| A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z | AA | AB | AC | AD | AE | AF | AG | AH | AI | AJ | AK | AL | AM | AN | AO | AP | AQ | AR | AS | AT | AU | AV | AW | AX | AY | AZ | BA | BB | BC | BD | BE | BF | BG | BH | BI | BJ | BK | BL | BM | BN | BO | BP | BQ | BR | BS | BT | BU | BV | BW | BX | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1 | Paper information | Algorithm Methodology | Other algorithms used | Information used | Offline/Online/UserStudy | Type of split | Evaluation methodology | Datasets/Source code | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
2 | Year | Acronym | Title | Authors | Paper URL | Venue | Conference/Journal | DBLP URL | Query IDs | Citation Count (in Scopus 14 abril 2021) | Citation Command | General Comments. Reason why it should be ignored etc | Representative Paper | Sim CF (using a structure similar to KNN) | Factorization | Probabilistic | Deep Learning | Social Graph/Link-analysis | Hybrid | Other | Optimization criterion (If required) | Clustering | Embedding(word2vec, graph embedding, metric embedding) | Geographical | Social | User/Item content information | Textual | Sequential | Temporal | Offline | Online | User Study | Temporal: Global-Temporal split with a global moment of split | Temporal: User-Temporal split for every user | Random Global: random split in a global manner | Random User: random split for every user | n-foldGlobal | n-foldUser | Other(e.g. different cities for training/test, user case study etc) | Filter users/items | Independent cities/Filtered regions. SPECIFICALLY STATED IN THE PAPER. If they refer those details to other paper it does not count, it needs to be specifically stated | Error/Ranking Metrics | Metrics | Test against classic Rec algorithms (Yes if Pop Is being tested as baseline) | Classic Non personalized baseline (Pop or random) | Classic personalized recommender (UB, IB, BPR, MFs, but not PMF. PMF is not classical. Classical MFs are SVDs, Non matrix factorization, o weighted matrix factorization) | Geographical baseline | Use validation subset | Split by Check-ins or locations/POIS | Cold start analysis? | Gowalla | Foursquare | Yelp | Other | Public dataset | Public statistics of the dataset?. Users, POIs, Check-ins | Prev/Post statistics | Public source code (???) | Source code: proposal | Source code: baselines | Source code: methodology (splitting, evaluation, etc.) | Dataset Info | ||||||||||||||
3 | 2011 | USG | Exploiting geographical influence for collaborative point-of-interest recommendation | Ye, M., Yin, P., Lee, W.-C., Lee, D.-L. | https://dl.acm.org/citation.cfm?id=2009962 | SIGIR | Conference | 1 | 784 | \cite{DBLP:conf/sigir/YeYLL11} | Yes | Yes(friends and preferences) | No | Yes(part of the hybrid approach, for modeling geographical influence, using power law distribution) | No | No | Yes(KNN, SocialKNN and Geographical) | No | No | No | Yes | Yes | No | No | No | No | Yes | No | No | No | No | No | Yes(10, 30 or 50% of the user locations sent to test) | No | No | No | No | Ranking | Precision and Recall | No(Random Walk, other combinations of USG) | No | Yes(UB CF) | Yes(USG with only the geographical part) | No | POIs | Yes(users with less than 5 interactions) | No | Yes | No | Whrrl | No | Not complete. For both Foursquare and Whrrl we do not know the number of Check-ins | Prev-No filtering | No | None. They crawled the data | |||||||||||||||||||||
4 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
5 | 2012 | GTS-FR | Followee recommendation in asymmetrical location-based social networks | Ying, J.J.-C., Lu, E.H.-C., Tseng, V.S. | https://dl.acm.org/citation.cfm?doid=2370216.2370431 | UbiComp | Conference | 1,3 | 11 | I think it is not on the scope. Recommending followees. IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
6 | 2012 | ---No-Acronym-- | A Study of Recommending Locations on Location-Based Social Network by Collaborative Filtering | Zhou, Dequan and Wang, Bin and Rahimi, Seyyed Mohammadreza and Wang, Xin | https://link.springer.com/chapter/10.1007/978-3-642-30353-1_22 | Canadian Conference on Artificial Intelligence | Conference | 3 | 0 | IGNORE. No model proposed. Experiments on PLSA, KNN with no proper model | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
7 | 2012 | LARS | LARS: A location-aware recommender system | Levandoski, J.J., Sarwat, M., Eldawy, A., Mokbel, M.F. | https://ieeexplore.ieee.org/document/6228105 | ICDE | Conference | 1,3 | 295 | \cite{DBLP:conf/icde/LevandoskiSEM12} | Yes | Yes | No | No | No | No | No | No | No | No | Yes | No | No | No | No | No | Yes | No(although they claim to use an online recommender) | No | No | No | Yes | No | No | No | No | Yes(Minnesota, USA for Foursquare, Minnesota for Synthetic) | None | Storage, Locality Loss | No(IB and variations of LARS) | No | Yes(IB) | Yes(LARS with only travel penalty) | No | Ratings, so POIs | No | No | Yes | No | Yes(Movielens, Synthetic) | No(only for Movielens) | Yes/Not complete for the previous stats | Both for Foursquare | No | None. They crawled the data from Foursquare | |||||||||||||||||||||
8 | 2012 | ---No-Acronym-- | Location-based and preference-aware recommendation using sparse geo-social networking data | Bao, J., Zheng, Y., Mokbel, M.F. | https://dl.acm.org/doi/10.1145/2424321.2424348 | SIGSPATIAL | Conference | 1 | 477 | \cite{DBLP:conf/gis/0003ZM12} | Yes | Yes | No | No | No | Yes(They claim to use HITS and so on, but in users and categories) | No | No | No | No | Yes(They define a spatial range, and I think this can be considered as geographical as the local experts are selected this way) | No(although they claim applying social information, they are not friends of target user, experts.) | Yes | No | No | No | Yes | No(although they claim to use an online recommender) | No | No | No | No | No | No | No | Yes(They select a target city and make recommendations to that target query city) | Yes(only users from New Jersey and users need to have at least 8 tips) | Yes(Los Angeles and NYC) | Ranking | Precision and Recall | No(MPC, LCF, PCF) | No | Yes(LCF is classic UB) | No | No | POIs | No | No | Yes | No | No | No | Not complete(number pois not stated) | Both | No | None. They crawled the data from Foursquare. Not complete | ||||||||||||||||||||
9 | 2012 | UPOI-Mine | Urban point-of-interest recommendation by mining user check-in behaviors | Ying, J.J.-C., Lu, E.H.-C., Kuo, W.-N., Tseng, V.S. | https://dl.acm.org/citation.cfm?id=2346507 | SIGKDD International Workshop on Urban Computing | Conference | 1,3 | 73 | \cite{DBLP:conf/kdd/YingLKT12} | Yes | Yes | No | No | No | No | Yes(Social, individual preferences and popularity in a regression model. I think it is hybrid) | No | No | No | Yes(Similarity based on distance) | Yes | Yes(preference in category) | No | No | No | Yes | No | No | No Information | No Information | No Information | No Information | No Information | No Information | No Information | No | No | Error/Ranking | MAE and NDCG | No(TrustWalker, USG) | No | Yes(UB-CF) | Yes(UB-CF, with geographical) | No | POIs. They aggregate | No | Yes | No | No | No | No | Yes | Prev-No filtering | No | None. They crawled the data from Gowalla | ||||||||||||||||||||
10 | 2012 | RW, Weighted-RW | A random walk around the city: New venue recommendation in location-based social networks | Noulas, A., Scellato, S., Lathia, N., Mascolo, C. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=6406279 | SocialCom/PASSAT | Conference | 1 | 139 | \cite{DBLP:conf/socialcom/NoulasSLM12} | Yes | No | No | Yes | No | Yes(Random Walk) | No | No | No | No | No | Yes | No | No | No | No | Yes | No | No | Yes(multiple training and test consisting in 30 consecutive days) | No | No | No | No | No | No | No | Yes(Austin, Boston, Dallas, Denver, London, Los Angeles, NYC, Paris, San Francisco, Seattle, Seoul) | Ranking | Precision, Recall, APR | Yes(Popularity, Random, Activity, Home, Social, KNN, PlaceNet, MF) | Yes(popularity and random) | Yes(UBKNN) | Yes(home distance) | No | Check-ins BUT the already items visited in train are removed | No | Yes | Yes | No | No | No | Yes | Prev | No | None. They crawled the data from Gowalla(12M Check-ins) and Forusquare (35M for Foursquare) | ||||||||||||||||||||
11 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
12 | 2013 | ---No-Acronym-- | Social recommendation in location-based social network using text mining | Feitosa, R.M., Labidi, S., Dos Santos, A.L.S., Santos, N. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=6498238 | ISMS | Conference | ??? | 1,3 | 5 | I think it is not on the scope. Not POI recommendation IGNORE | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
13 | 2013 | ---No-Acronym-- | New perspectives for recommendations in location-based social networks: Time, privacy and explainability | Kefalas, P., Symeonidis, P., Manolopoulos, Y. | https://dl.acm.org/citation.cfm?id=2536202 | MEDES | Conference | 1 | 11 | I think it is not on the scope. It does not propose any algorithm. Its like a mini-survey on LBSNs. IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
14 | 2013 | CRTCF | Cross-region collaborative filtering for new point-of-interest recommendation | Zheng, N., Jin, X., Li, L. | https://dl.acm.org/doi/abs/10.1145/2487788.2487804 | WWW | Conference | 1 | 13 | \cite{DBLP:conf/www/ZhengJL13} | Yes | Yes(use LDA for topic modeling) | Yes(use LDA for topic modeling) | No | No | No | No | No | No | No | Yes | Yes | No | No | No | Yes | No | No | No | No | No | No | No | No | Yes(Major region as training set, rest to test set for every user) | No | Yes(New York City: Queens, Bronx, Brooklyn) | Ranking | Recall | No(Naive CF) | No | Yes | No | No | POIs(besides, they sum the interactions, at least in training) | No | Yes | No | No | No | No(They refer to Friendship and mobility: user movement in location-based social networks 2011) | Not complete(not stated the number of POIs) | Prev-No filtering | No | Gowalla: They refer to Friendship and mobility: user movement in location-based social networks 2011 | |||||||||||||||||||||
15 | 2013 | GMM, GA-GMM | Capturing geographical influence in POI recommendations | Zhao, S., King, I., Lyu, M.R. | https://link.springer.com/chapter/10.1007%2F978-3-642-42042-9_66 | ICONIP | Conference | 1,3 | 25 | \cite{DBLP:conf/iconip/ZhaoKL13} | No | No | Yes(Gaussian distribution) | No | No | No | Yes(for The GA-GMM uses a genetic algorithm model) | Expectation maximization | No | No | Yes | No | No | No | No | No | Yes | No | No | Yes/(They talk about sequence. 10% test rest to train and to redundant) | No | No | No | No | No | Yes(removed locations with less than 10 visits) | No | Ranking | Precision and Recall | No(GM, MGM, GMM, GA-GMM) | No | No | Yes(MGM) | No | Check-ins | No | Yes | No | No | No | No | Not complete(number of Check-ins not stated) | Post | No | Gowalla: None. They crawled the data from Gowalla | |||||||||||||||||||||
16 | 2013 | PCR (PCR is just for Category recommendation), PCLR | Location recommendation based on periodicity of human activities and location categories | Rahimi, S.M., Wang, X. | https://link.springer.com/chapter/10.1007/978-3-642-37456-2_32 | PAKDD | Conference | 1 | 23 | \cite{DBLP:conf/pakdd/RahimiW13} | No | No | Yes | No | No | Yes(category, geographical, temporal) | No | No | No | Yes | No | Yes | No | No | Yes | Yes | No | No | No | No | No | Yes(Random 1 Check-in for each user to test. Repeated 5 times. Leave one out) | No | No | No | No | Ranking | Precision and Recall | No(PMM+c, USG+c) | No | No | Yes(USG) | No | Check-ins | No | Yes | No | No | No | No | Yes | Prev-No filtering | No | Gowalla: They refer to: A Study of Recommending Locations on Location-Based Social Network by Collaborative Filtering 2012 | ||||||||||||||||||||||
17 | 2013 | GeoSocialRec | GeoSocialRec: Explaining recommendations in location-based social networks | Symeonidis, P., Krinis, A., Manolopoulos, Y. | https://link.springer.com/chapter/10.1007/978-3-642-40683-6_7 | ADBIS | Conference | 1 | 11 | \cite{DBLP:conf/adbis/SymeonidisKM13} | No | Yes | No | No | No | No | Higher Order Singular Value Decomposition (HOSVD) | No | No | No(Used in the friendlink algorithm, but not for POI recommendation) | No(only for the part of friend recommendation) | No | No | No | No | Yes | No(although they claim to use an online recommender) | Yes | No | No | No | No | No | Yes(4-fold cross validation for every user) | No | No | Ranking | Precision and Recall | No | No | No | No | No | Check-ins | No | No | No | No | GeoSocialRec and UCLAF | No(They refer to http://delab.csd.auth.gr/~symeon but i do not find there the dataset and they also refer to paper Collaborative location and activity recommendations with GPS history data 2010 | Yes | Prev-No filtering | No | Other dataset: UCLAF dataset used in: Collaborative filtering meets mobile recommendation: A user-centered approach. GeoSocialRec is extracted from http://delab.csd.auth.gr/~symeon | ||||||||||||||||||||||
18 | 2013 | LBSMF | A sentiment-enhanced personalized location recommendation system | Yang, D., Zhang, D., Yu, Z., Wang, Z. | https://dl.acm.org/citation.cfm?id=2481505 | HT | Conference | 1 | 158 | \cite{DBLP:conf/ht/YangZYW13} | Yes(They compute similarities of their friends) | Yes(Probabilistic MF) | Yes(Probabilistic MF) | No | No | No | Gradient descent | No | No | No | Yes | No | Yes(language processing) | No | No | Yes | No | No | No | No | Yes(90 or 80 of Check-ins to training) | No | No | No | Yes(1 Check-ins per week for the users and other conflict data) | Yes(New York and London) | Error | MAE and RMSE | No(CF, PMF, SosialMF, SRMF, LBSMF) | No | Yes(CF and MF) | No | No | Check-ins | No | No | Yes | No | No | No | Yes | Post | No | Foursquare: None. They crawled the data from Foursquare | ||||||||||||||||||||||
19 | 2013 | ---No-Acronym-- | A HITS-based POI recommendation algorithm for location-based social networks | Long, X., Joshi, J. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=6785770 | ASONAM | Conference | 1,3 | 23 | \cite{DBLP:conf/asunam/LongJ13} | No | No | No | No | Yes(HITS) | No | No | No | No | No | Yes | No | No | No | No | Yes | No | No | No | No | Yes(80% training 20% test) | No | No | No | Yes(removed items and users with few Check-ins but no more information said) | Yes(Pittsburgh area) | Ranking | Precision and Recall | No(rwr) | No | No | No | No | Check-ins | No | No | Yes | No | No | No | Yes | Prev | No | Foursquare: None. They crawled the data from Foursquare | ||||||||||||||||||||||
20 | 2013 | TL-PMF | Point-of-interest recommendation in location based social networks with topic and location awareness | Liu, B., Xiong, H. | https://epubs.siam.org/doi/10.1137/1.9781611972832.44 | SIAM | Conference | 1 | 140 | \cite{DBLP:conf/sdm/LiuX13} | No | Yes | Yes(It is probabilistic matrix factorization) | No | No | No | Gradient descent | No | No | No(only for recommendation, the use it to filter a range) | No | Yes(tags/categories) | Yes(textual) | No | No | Yes | No | No | No | No | Yes(80% o the data to training rest to test) | No | No | No | Yes(users with less than 6 Check-ins removed and POIs with minimum 10 tags) | Yes(Different regions. California (CA), Arizona (AZ), Texas (TX), Florida (FL), Chicago area (IL), Washington DC (DC), Boston area (MA) and New York area (NY)) | Error | RMSE | No(PMF) | No | No(PMF) | No | No | POIs (More less stated that the training matrix is built by aggregating) | No | No | Yes | No | No | No(They refer to this paper: Toward traffic-driven location-based web search 2011) | Yes | Post | No | Foursquare: They refer to this paper: Toward traffic-driven location-based web search 2011 | ||||||||||||||||||||||
21 | 2013 | CLW | Evaluation of social, geography, location effects for point-of-interest recommendation | Cheng, N.-H., Chang, C.-H. | https://ieeexplore.ieee.org/document/6753998 | ICDM | Conference | 1,3 | 1 | \cite{DBLP:conf/icdm/ChengC13} | Yes | No | No | No | No | Yes(User UBKNN, IBKNN, Social influence) | No | No | No | No | Yes | No | No | No | Yes | Yes | No | No | No | No | No | Yes(70% users records as training 30% to test) | No | No | No | Yes(New York and San Francisco) | Ranking | Precision and Recall | No(LRALL) | No | Yes(U is UB) | Yes(Figure 9 says something about geographical) | No | Check-ins | Yes(users with less than 5 Check-ins) | Yes | No | No | No | No(They refer to this paper: Learning Collaborative Filtering and Its Application to People to People Recommendation in Social Networks 2010) | Yes | Prev-No filtering | No | Gowalla: They refer to this paper: Learning Collaborative Filtering and Its Application to People to People Recommendation in Social Networks | ||||||||||||||||||||||
22 | 2013 | ---No-Acronym-- | Personalized point-of-interest recommendation by mining users' preference transition | Liu, X., Liu, Y., Aberer, K., Miao, C. | https://dl.acm.org/doi/10.1145/2505515.2505639 | CIKM | Conference | 1 | 162 | \cite{DBLP:conf/cikm/LiuLAM13} | Yes | No | Yes | Yes(power law for modeling geographic information) | No | No | Yes(geographical + MF) | No optimization algorithm stated | Yes | No | Yes | No | Yes | No | Yes(transition probabilities) | Yes(for building sequences) | Yes | No | No | No | Yes(70% training, 30% test for every user) | No | No | No | No | No | Yes(Austin, Chicago, Houston, Los Angeles and San Francisco) | Ranking | Precision and Recall | No(MF, GeoCF, MGMMF, Markov, MF) | No | Yes(BasicMF) | Yes(GeoCF, MGMM) | No | Check-ins | No | Yes | No | No | No | No | Yes | Prev-No filtering | No | Gowalla: None. They crawled the data from Gowalla | |||||||||||||||||||||
23 | 2013 | UPS-CF | Location recommendation for out-of-town users in location-based social networks | Ference, G., Ye, M., Lee, W.-C. | https://dl.acm.org/doi/10.1145/2505515.2505637 | CIKM | Conference | 1 | 108 | \cite{DBLP:conf/cikm/FerenceYL13} | Yes | No | No | No | No | No | No | No | No | No(I think only for recommendation, not in the model) | Yes | No | No | No | No | Yes | No | No | No | No | No | Yes(randomly remove 1 Check-in for every user. Leave one out) | No | No | No | No | Ranking | Precision | Yes(Pop, CL, UB, UP, US, UPS) | Yes(Pop) | Yes(UBCF) | Yes(Closest locations | No(Cross-validation) | Check-ins | Yes (at most 2 locations visited users) | Yes | Yes | No | No | No | Not complete(number of Check-ins not stated) | Prev-No filtering | No | Gowalla, Foursquare: None. They crawled the data from Gowalla and Foursquare | ||||||||||||||||||||||
24 | 2013 | GT-BNMF | Learning geographical preferences for point-of-interest recommendation | Liu, B., Fu, Y., Yao, Z., Xiong, H. | https://dl.acm.org/doi/10.1145/2487575.2487673 | SIGKDD | Conference | 1 | 309 | \cite{DBLP:conf/kdd/LiuFYX13} | Yes | No | Yes(probabilistic MF) | Yes(probabilistic MF) | No | No | No | Expectation maximization | Yes | No | Yes | No | No | Yes(textual information) | No | No | Yes | No | No | No | No | Yes(80% training - 20% test) | No | No | No | No | Yes(USA) | Ranking | Precision, Recall, nPrecision and nRecall | No(SVD, NMF, PMF) | No | Yes(MF, SVD) | No | No | Check-ins | No | No | Yes | No | No | No(They refer to Point-of-interest recommendation in location based social networks with topic and location awareness 2013 and Toward traffic-driven location-based web search 2011) | Yes | Prev-No filtering | No | Foursquare: They refer to Point-of-interest recommendation in location based social networks with topic and location awareness 2013 and Toward traffic-driven location-based web search 2011 | |||||||||||||||||||||
25 | 2013 | FPMC-LR | Where you like to go next: Successive point-of-interest recommendation | Cheng, C., Yang, H., Lyu, M.R., King, I. | https://dl.acm.org/doi/10.5555/2540128.2540504 | IJCAI | Conference | 1 | 329 | \cite{DBLP:conf/ijcai/ChengYLK13} | Next POI recommendation | Yes | No | Yes | Yes(FPMC, BPR) | No | No | No | Sequential BPR | No | No | Yes | No | No | No | Yes | No | Yes | No | No | Yes(check-ins in the last time slot are test data and rest are for train) | No | No | No | No | No | Yes(users need to have checked-in at least 120 times and each location should have visited at least 5 times) | No | Ranking | Precision and Recall | No(PMF, PTF, FPMC) | No | No(PMF) | No | No(validation for one parameter in their approach) | Check-ins | No | Yes | Yes | No | No | Refer to other papers: Exploring millions of foot- prints in location sharing services (2011) and Fused matrix factorization with geographical and social influence in location-based social networks (2012) | Yes | Post | No | Gowalla and Foursquare: Refer to other papers: Exploring millions of foot- prints in location sharing services (2011) and Fused matrix factorization with geographical and social influence in location-based social networks (2012) | ||||||||||||||||||||
26 | 2013 | LRT | Exploring temporal effects for location recommendation on location-based social networks | Gao, H., Tang, J., Hu, X., Liu, H. | https://dl.acm.org/doi/10.1145/2507157.2507182 | RecSys | Conference | 1,3 | 338 | \cite{DBLP:conf/recsys/GaoTHL13} | Yes | No | Yes | No | No | No | No | No optimization algorithm stated | No | No | No | No | No | No | No | Yes | Yes | No | No | No | No | No | Yes(20% and 40% of random POIs for every user to test) | No | No | Yes(users with at least 10 different POIs and POIs that have at least 2 different users who have cheked-in) | No(Although The majority of the Check-ins are from USA there is not proof that maybe other chekcins are around the world) | Ranking | Precision and Recall | No(USG, R-LRT, NMF) | No | Yes(CF, NMF) | No | No(parameter optimization by cross-validation) | POIs | No | No | Yes | No | No | Yes(http://www.public.asu.edu/~hgao16/dataset.html) but the url does not work anymore | Yes | Post | No | Foursquare: http://www.public.asu.edu/~hgao16/dataset.html. URL does not work | |||||||||||||||||||||
27 | 2013 | LFBCA | Location recommendation in location-based social networks using user check-in data | Wang, H., Terrovitis, M., Mamoulis, N. | https://dl.acm.org/citation.cfm?id=2525357 | SIGSPATIAL | Conference | 1,3 | 140 | \cite{DBLP:conf/gis/WangTM13} | No | No | No | No | Yes | No | No | No | No | Yes(Although it does not build a model based on geographical information, it uses the distance of the POIs in it) | Yes | No | No | No | No | Yes | No | No | Yes(Weird specification) | No | No | No | No | No | Yes(in all the experiments we consider only active users, i.e., users who have at least one new visit in the testing period) | No | Ranking | Precision, Recall, Utility | No(UBKNN, LocCF, FriendCF, LocNN, RWR) | No | Yes(UB, IB) | Yes(LocNN) | No | Check-ins | No | Yes | No | No | Brightkite | Yes(http://snap.stanford.edu/index.html) | Yes | Prev | No | Gowalla, Brightkite: http://snap.stanford.edu/index.html. They also refer to this paper Friendship and mobility: user movement in location-based social networks 2011 | ||||||||||||||||||||||
28 | 2013 | iGSLR | iGSLR: Personalized geo-social location recommendation: A kernel density estimation approach | Zhang, J.-D., Chow, C.-Y. | https://dl.acm.org/citation.cfm?id=2525339 | SIGSPATIAL | Conference | 1 | 150 | \cite{DBLP:conf/gis/ZhangC13} | Yes | No | Yes (KDE can be considered as probabilistic) | No | No | Yes (Social, KDE) | No | No | No | Yes | Yes | No | No | No | No | Yes | No | No | Yes(90% for training 10% for test) | No | No | No | No | No | No | No | Ranking | Precision and Recall | No(UBKNN, SCF, UPOI-Mine) | No | Yes(UB) | Yes(geographical CF) | No | Check-ins | Yes(analysis with users with 1, 2 ,3 ,4 ,5 ,6 ,7 ,8, 9 and 10 visited locations in the training set) | Yes | Yes | No | No | Yes(http://www.public.asu.edu/~hgao16/dataset.html) but the url does not work anymore and (https://snap.stanford.edu/data/loc-gowalla.html) for Gowalla | Yes | Prev-No filtering | No | Gowalla, Foursquare: http://www.public.asu.edu/~hgao16/dataset.html but the url does not work anymore and (https://snap.stanford.edu/data/loc-gowalla.html) for Gowalla. They also refer to this paper Friendship and mobility: user movement in location-based social networks 2011 and gscorr: Modeling geo-social correlations for new check-ins on location-based social networks 2012 | ||||||||||||||||||||||
29 | 2013 | UTE, SE, UTE+SE | Time-aware point-of-interest recommendation | Yuan, Q., Cong, G., Ma, Z., Sun, A., Magnenat-Thalmann, N. | https://dl.acm.org/citation.cfm?id=2484030 | SIGIR | Conference | 1 | 535 | \cite{DBLP:conf/sigir/YuanCMSM13} | Time Aware POI recommendation. | Yes | Yes | No | Yes(They use power-law distribution) | No | No | Yes(Temporal and spatial influences) | No | No | No | Yes | No | No | No | No | Yes | Yes | No | No | No | No | No | Yes(12.5% of random POIs to validation, 25% to test) | No | No | Yes(remove users and items with less that 5 check-ins) | Yes(California, Nevada for Gowalla, Singapore for Foursquare) | Ranking | Precision and Recall | No (UB-KNN, UTF, UT, UTE, SB, S, SE...) | No | Yes(UB) | Yes(Spatial influence baseline) | Yes | POIs | No | Yes | Yes | No | No | Yes(https://www.ntu.edu.sg/home/gaocong/datacode.htm) | Yes | Post | No | Gowalla and Foursquare: They refer to Friendship and mobility: user movement in location-based social networks 2011 for Gowalla. Foursquare | ||||||||||||||||||||
30 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
31 | 2014 | ---No-Acronym-- | Enhancing a location-based recommendation system by enrichment with structured data from the web | Schmachtenberg, M., Strufe, T., Paulheim, H. | https://dl.acm.org/doi/abs/10.1145/2611040.2611080 | WIMS | Conference | 1 | 13 | They do not propose a model, just to extend the data (similar to cross-domain?) --> they do not propose a recommendation algorithm, but how to extend the item data with semantic information. IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
32 | 2014 | PATR | Trip recommendation with multiple user constraints by integrating point-of-interests and travel packages | Fang, S.H., Lu, E.H.C., Tseng, V.S. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=6916901 | MDM | Conference | 1 | 12 | Its trip recommendation, not POI. IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
33 | 2014 | ---No-Acronym-- | Personalized location recommendation on location-based social networks | Gao, H., Tang, J., Liu, H. | https://dl.acm.org/citation.cfm?id=2645710.2645776 | ICCC | Conference | 1,3 | 14 | It is a tutorial. IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
34 | 2014 | Recommender Systems for Location-Based Social Networks | Symeonidis, Panagiotis and Ntempos, Dimitrios and Manolopoulos, Yannis | https://link.springer.com/book/10.1007%2F978-1-4939-0286-6 | 3 | - | IGNORE. Its a book | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
35 | 2014 | ---No-Acronym-- | Locations recommendation based on check-in data from Location-Based Social Network | Jiang, D., Guo, X., Gao, Y., Liu, J., Li, H., Cheng, J. | https://ieeexplore.ieee.org/document/6950814 | Geoinformatics | Conference | 1 | 13 | \cite{DBLP:conf/geoinformatics/JiangGGLLC14} | Not sure if it is on the scope. Propose 4 methods with no standard evaluation. IGNORE | Yes | No | No | No | No | No | No | No | No | Yes | No | Yes | No | No | No | Yes | No | No | No Information | No Information | No Information | No Information | No Information | No Information | No | Yes(China, removed the data out of China) | None | None | No | No | No | No | No | No Information | No | No | No | No | Yes(Datatong) | Yes(www.datatang.com/data/43896 but the url does not work) | Not complete(number of POIs not stated) | Prev-No filtering | No | Datatong: Url does not work | |||||||||||||||||||||
36 | 2014 | CGAR | Collaborative group-activity recommendation in location-based social networks | Purushotham, S., Jay Kuo, C.-C., Shahabdeen, J., Nachman, L. | https://dl.acm.org/citation.cfm?id=2676442 | GeoCrowd/SIGSPATIAL | Conference | 1,3 | 16 | \cite{DBLP:conf/gis/PurushothamKSN14} | It is group oriented. IGNORE | No | Yes | Yes | No | No | Yes(Fuse MF and Latent Dirichlet Allocation) | Gradient descent. | Yes(for creating the groups) | No | Yes | No | No | No | No | Yes | Yes | No | No | No | No | Yes(80% training 5% validation 15% test) | No | No | No | Yes(only groups with at least 10 Check-ins) | No | Error/Ranking | Recall, Average Accuracy and RMSE | No(MF, CTR, Aggregation methods for group recommendation) | No | Yes(MF) | No | Yes | Check-ins | No (they discuss about it but i do not see any specific comparison) | Yes | No | No | No | Yes(http://snap.stanford.edu/data/#locnet) | Yes | Post | No | Gowalla: They refer to http://snap.stanford.edu/data/#locnet | |||||||||||||||||||||
37 | 2014 | LARS* | LARS*: An efficient and scalable location-aware recommender system | Sarwat, M., Levandoski, J.J., Eldawy, A., Mokbel, M.F. | https://ieeexplore.ieee.org/document/6427747 | TKDE | Journal | 1,3 | 118 | \cite{DBLP:journals/tkde/SarwatLEM14} | Basically it is an extension of 2012 LARS. IGNORE. The same as \cite{DBLP:conf/icde/LevandoskiSEM12} | Yes | No | No | No | No | No | No | No | No | Yes | No | No | No | No | No | Yes | No(although they claim to use an online recommender) | No | No | No | Yes(CAREFUL. IT IS NOT SPECIFICALLY STATED THAT IT IS RANDOM. 80% of ratings to training rest to test) | No | No | No | Yes(they take a small subset from the original Foursquare dataset crawled) | Yes(United States for FOursquare and Minnesota for synthetic) | None | Storage, Response time, locality | No(LARS) | No | No | Yes(LARS) | No | Check-ins | No | No | Yes | No | Yes(Movielens, Synthetic) | No(only for movielens http://www.movielens.org) | Yes | Post | No | Foursquare: None. They crawled the data from Foursquare | |||||||||||||||||||||
38 | 2014 | Ricochet | Ricochet: Context and complementarity-aware, Ontology-based POIs recommender system | Lu, C., Laublet, P., Stankovic, M. | http://ceur-ws.org/Vol-1165/salad2014-3.pdf | Second Workshop on Services and Applications over Linked APIs/ESWC | Conference | 1 | 1 | \cite{DBLP:conf/esws/LuLS14} | Not sure if it is on the scope | No | No | No | No | No | Yes(Weather, complementary and time of day) | No | No | No | No | No | Yes | No | No | Yes | No | No | Yes | No | No | No | No | No | No | Yes(User study, it seems user study) | No | Yes(Paris) | Ranking | Precision, Recall and NDCG | No(only 2 versions of the proposed algorithm) | No | No | No | No | No Information | No | No | No | Yes(they use it but they do not recommend it) | No | No | No | None | No(although they give some urls of the things they have used: http://www.w3.org/TR/owl-ref/#OWLFull and http://sep- age.com/ontology/OntoPOI_Ricochet.owl) | Yelp: None | ||||||||||||||||||||
39 | 2014 | MO-CI, MO-CFI, MO-CFIH, MO-CH) | Multi-objective optimization based location and social network aware recommendation | Abdel-Fatao, H., Li, J., Liu, J. | https://ieeexplore.ieee.org/document/7014569 | CollaborateCom | Conference | 1 | 8 | \cite{DBLP:conf/colcom/OzsoyPA14} | Yes | No | No | No | No | No | No | No | No | Yes(filter users by the hometown) | Yes(filter users in the friends) | No | No | No | No | Yes | No | No | Yes(January 2011 for train and February 2011 for test) | No | No | No | No | No | Yes(from the original dataset they only take into account January 2011) | No | Ranking | Precision, HitRate, Coverage and NDCG | No(CF-C, CF-F, CF-I, CF-H) | No | Yes(CF) | Yes(CF with hometown) | No | Check-ins (Although they define a process to group the interactions, from implicit to explicit) | No (they discuss about it) | No | Yes | No | No | No(refers to this URL http://www.public.asu.edu/∼hgao16/dataset.html but it does not work) | Yes | Post | No | Foursquare: Refers to Location-based social network data repository 2014 with url http://www.public.asu.edu/∼hgao16/dataset.html but it does not work | ||||||||||||||||||||||
40 | 2014 | PNN | A Location Recommender System for Location-Based Social Networks | Kosmides, P., Remoundou, C., Demestichas, K., Loumiotis, I., Adamopoulou, E., Theologou, M. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7046197 | MCSI | Conference | ??? | 1,3 | 8 | \cite{Kosmides2014} | No | No | Yes(Probabilistic Neural Network) | Yes(Probabilistic Neural Network) | No | No | No | No | No | Yes(input of the neural network, latitude and longitude) | Yes(input of the neural network, friend of the user) | No | No | No | Yes(day and hour as input for the algorithm) | Yes | No | No | No | No | No | No | Yes(10 fold cross-validation, altough not stated as CC) | No | No | No(They say they use the same dataset than the one in LARS, and in that paper they DO filter) | None | Misclassification Percentage | No(NN, SVM) | No | No | No | No | No Information | No | No | Yes | No | No | No(They refer to LARS: A Location-Aware Recommender System 2012) | No | None | No | Foursquare: They refer to LARS: A Location-Aware Recommender System 2012 | |||||||||||||||||||||
41 | 2014 | BPTFSLR | Personalized recommendations of locally interesting venues to tourists via cross-region community matching | Zhao, Y.-L., Nie, L., Wang, X., Chua, T.-S. | https://dl.acm.org/citation.cfm?id=2532439 | TIST | Journal | 1 | 49 | \cite{DBLP:journals/tist/ZhaoNWC14} | No(They define venue similarities but in a content way) | Yes | Yes(Probabilistic MF) | No | No | No | Gradient descent. Not sure | Yes | No | No(They state local communities but to evaluate) | Yes(Part of the input, social relation matrix) | No | No | No | Yes(user poi matrix matrix) | Yes | No | No | Yes(August-October 2012 as training rest to test) | No | No | No | No | No | Yes(removed Check-ins from users who have performed more than 10 check-ins per minute and removed check-ins faster than 1000km/h) | Yes(NewYork, Singapore, Chicago, London) | Ranking | Recall and MAP | Yes(Pop, UCF, SVD, NMF, BPMF, BPTF) | Yes(Pop) | Yes(CF and MF) | No | No | Check-ins | No | No | Yes | No | No | No | Yes | Post | No | Foursquare: None. They crawled the data from Foursquare | ||||||||||||||||||||||
42 | 2014 | ST | Social Topic Modeling for Point-of-Interest Recommendation in Location-Based Social Networks | Hu, B., Ester, M. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7023411 | ICDM | Conference | 1,3 | 40 | \cite{DBLP:conf/icdm/HuE14} | No | Yes(Topic modeling similar to LDA, but it is NOT LDA) | Yes(Topic modeling similar to LDA, but it is NOT LDA) | No | No | No | No | No | No | No | Yes(user friends) | Yes(they consider tags and documents as words) | Yes(reviews for topic modeling) | No | No | Yes | No | No | No | No | No | Yes(70% training. 30% test for each user) | No | No | No | Yes(Yes for Yelp but no for Foursqaure) | Ranking | Recall | Yes(Pop, PMF, LDA, PMFSR, SLDA, STT) | Yes(Pop) | No(PMF) | No | No | Check-ins | No | No | Yes | Yes | No | Yes(http://www.sfu.ca/˜boh but the url does not work) | Yes | Prev-No filtering | Yes(http://www.sfu.ca/˜boh but the url does not work) | No | No | No | Foursquare and Yelp. They refer to http://www.sfu.ca/˜boh but URL does not work | |||||||||||||||||||
43 | 2014 | DGM | A personalized geographic-based diffusion model for location recommendations in LBSN | Nunes, I., Marinho, L. | https://ieeexplore.ieee.org/document/7000172 | LA-WEB | Conference | 1,3 | 5 | \cite{DBLP:conf/la-web/NunesM14} | Yes | No | No | No | Yes | No | No | No | No | Yes | No | No | No | No | No | Yes | No | No | No | No | No | Yes(for each user, removed 10% of their Check-ins locations to test and rest to train. Repeated) | No | No | Yes(users with more than 10 locations visited and locations with more than 10 users that have visited them) | Yes(New York, Los Angeles and Chicago for Foursquare and Austin and Stockholm for Gowalla) | Ranking | Precision and Recall | No(FMFMGM, UG) | No | Yes(UG is UB) | Yes(FMFMGM) | No | Locations | No | Yes | Yes | No | No | Yes(http://snap.stanford.edu/data/loc-gowalla.html for Gowalla, and Foursquare http://infolab.tamu.edu/data/) | Yes | Post | No | Foursquare and Gowalla: http://snap.stanford.edu/data/loc-gowalla.html for Gowalla and http://infolab.tamu.edu/data/ for Foursquare | ||||||||||||||||||||||
44 | 2014 | GPLR, GPUR | Group-based personalized location recommendation on social networks | Wang, H., Li, G., Feng, J. | https://link.springer.com/chapter/10.1007/978-3-319-11116-2_7 | APWeb | Conference | 1 | 12 | \cite{DBLP:conf/apweb/WangLF14} | Yes | No | No | No | No | No | No | Yes | No | No(they say the use distance but i do not find anything) | Yes | Yes(TF-IDF categories) | No | No | No | Yes | No(although they claim to use an online recommender) | No | No | No | No | Yes(divided the location histories of a user into two parts) | No | No | No | Yes (New York and Los Angeles) | Ranking | Precision and Recall | No(LCF, UCF, PA) | No | Yes(UB and LCF) | No | No(cross-validation) | Check-ins (but for the CF component they take into account the number of times the user visited every location) | No (they discuss about it but i do not see any specific comparison) | No | Yes | No | No | No(they refer to paper Location-based and preference-aware recommendation using sparse geo-social networking data 2012) | Not complete(number of POIs not stated) | Prev-No filtering | No | Foursquare: For Foursquare: Location-based and preference-aware recommendation using sparse geo-social networking data 2012. It is the same as \cite{DBLP:conf/www/LiWW020} and DBLP:conf/wsdm/LiW020. Weird, more users than tips | ||||||||||||||||||||||
45 | 2014 | ITF | Location-based recommendation using incremental tensor factorization model | Zou, B., Zou, B., Li, C., Li, C., Tan, L., Tan, L., Chen, H., Chen, H. | https://link.springer.com/chapter/10.1007%2F978-3-319-14717-8_18 | ADMA | Conference | 1 | 2 | \cite{DBLP:conf/adma/ZouLTC14} | More oriented for events | No | Yes | No | No | No | No | Higher Order Singular Value Decomposition (HOSVD) | No | No | Yes | No | No | No | No | No | Yes | No(although they claim to use an online recommender) | No | No | No | Yes(50% training rest to incremental training of 10%) | No | No | No | No | No | Ranking | Precision and Recall | No(MF, TF) | No | Yes(MF) | No | No | Check-ins | No | No | No | No | DoubanEvent | Yes(http://net.pku.edu.cn/daim/yinhongzhi/index.html, although the URL does not work) | Yes | Prev-No filtering | No | DoubanEvent: They refer to this url but it doues not work http://net.pku.edu.cn/daim/yinhongzhi/index.html | |||||||||||||||||||||
46 | 2014 | UPOI-Walk | Mining user check-in behavior with a random walk for urban point-of-interest recommendations | Ying, J.J.-C., Kuo, W.-N., Tseng, V.S., Lu, E.H.-C. | https://dl.acm.org/citation.cfm?id=2523068 | TIST | Journal | 1,3 | 54 | \cite{DBLP:journals/tist/YingKTL14} | Yes | No | No | No | No | Yes(HITS based Random Walk) | No | No | No | No | Yes(distance) | Yes(friends) | Yes(category) | No | No | No | Yes | No(although they claim to use an online recommender) | No | No Information | No Information | No Information | No Information | No Information | No Information | No | Yes(They say Gowalla is only for New York?. No info for EveryTrail) | Error/Ranking | NDCG and MAE | No (USG, OtherCF, TrustWalker,HIST,UPOI-Mine) | No | Yes(CF) | Yes(UPOI-Mine) | No | No Information | No | Yes | No | No | EveryTrail | No | Yes | Prev-No filtering | No | EveryTrail and Gowalla: no further details | |||||||||||||||||||||
47 | 2014 | GTAG | Graph-based point-of-interest recommendation with geographical and temporal influences | Yuan, Q., Cong, G., Sun, A. | https://dl.acm.org/citation.cfm?id=2661983 | CIKM | Conference | 1 | 165 | \cite{DBLP:conf/cikm/YuanCS14} | Time Aware POI recommendation. | Yes | No | No | No | No | Yes | No | No | No | No | Yes | No | No | No | No | Yes | Yes | No | No | No | No | No | Yes(70% training, 10% tune and 20% test FOR EVERY USER) | No | No | Yes(users and items with at least 5 POIs, or POIs that have at least 5 users) | Yes(They refer to Time-aware point-of-interest recommendation 2013 (UTESE) and they say it is Singapore) | Ranking | Precision and Recall | No(UB-KNN, UTF, UTE, UTE-SE, MS-IPF, LRT, Other versions of the approach) | No | Yes(CF) | Yes(UG uses geographical information) | Yes | POIs | No | Yes | Yes | No | No | Yes(Same dataset as in Time-aware point-of-interest recommendation 2013) | Yes | Post | No | Gowalla and Foursquare: They refer to Time-aware point-of-interest recommendation 2013 | ||||||||||||||||||||
48 | 2014 | GeoMF | GeoMF: Joint geographical modeling and matrix factorization for point-of-interest recommendation | Lian, D., Zhao, C., Xie, X., Sun, G., Chen, E., Rui, Y. | https://dl.acm.org/doi/10.1145/2623330.2623638 | SIGKDD | Conference | 1,3 | 377 | \cite{DBLP:conf/kdd/LianZXSCR14} | Yes | No | Yes | No | No | No | No | Alternate least squares | No | No | Yes | No | No | No | No | No | Yes | No | No | No | No | No | Yes(70% for training 30% for test) | No | No | Yes(each POI visited by at least 2 users and each user need to visit at least 10 different POIs) | No | Ranking | Precision and Recall | No(UCF, MF, MF-Freq, B-NMF, WMF-B) | No | Yes(CF and MF) | No | No(Cross-validation) | POIs | No | No | No | No | Jiepang (chinese LBSN similar to Foursquare) | No | Not complete(number of Check-ins not stated) | Post | No | Jiepang: crawled from there | |||||||||||||||||||||
49 | 2014 | IRenMF | Exploiting geographical neighborhood characteristics for location recommendation | Liu, Y., Wei, W., Sun, A., Miao, C. | https://dl.acm.org/doi/10.1145/2661829.2662002 | CIKM | Conference | 1,3 | 190 | \cite{DBLP:conf/cikm/LiuWSM14} | Yes | No | Yes | No | No | No | No | Accelerated proximal gradient (APG) | Yes | No | Yes | No | No | No | No | No | Yes | No | No | No | No | No | Yes(60% train, 10% validation and 20% test for every user) | No | No | No | Yes(Berlin, London, Chicago, and San Francisco) | Ranking | Precision and Recall | No(UCF, ICF,WRMF, BPRMF, GeoCf, MGMMF, InMF) | No | Yes(CF) | Yes(GeoCF) | Yes | POIs | No | Yes(Facebook subset) | No | No | No | No | Not complete(number of Check-ins not stated) | Post | No | Gowalla: They crawled from Gowalla | |||||||||||||||||||||
50 | 2014 | LORE | LORE: Exploiting sequential influence for location recommendations | Zhang, J.-D., Chow, C.-Y., Li, Y. | https://dl.acm.org/citation.cfm?id=2666400 | SIGSPATIAL | Conference | 1 | 137 | \cite{DBLP:conf/gis/ZhangCL14} | Yes | Yes | No | Yes(KDE) | No | No | Yes(Social, KDE, Sequential) | No | No | No | Yes | Yes | No | No | Yes | No | Yes | No | No | Yes(50% train and 50% test by Check-ins) | No | No | No | No | No | No | No | Ranking | Precision and Recall | No(FMC, AMC, iGSLR, GS2D, FM+GS2D) | No | No | Yes(igslr + GS2D) | No | Check-ins | No | Yes | Yes | No | No | Yes(http://www.public.asu.edu/~hgao16/dataset.html) but this URL for Foursquare does not work and (https://snap.stanford.edu/data/loc-gowalla.html) for Gowalla | Yes | Prev-No filtering | No | Gowalla and Foursquare: They refer to http://www.public. asu.edu/~hgao16/Publications.html for Foursquare (url does not work) and http://snap. stanford.edu/data/loc-gowalla.html for Gowalla | |||||||||||||||||||||
51 | 2014 | sPCLR | Probabilistic category-based location recommendation utilizing temporal influence and geographical influence | Zhou, D., Wang, X. | https://ieeexplore.ieee.org/document/7058061 | DSAA | Conference | 1 | 8 | \cite{DBLP:conf/dsaa/ZhouW14} | Yes | No | No(They discuss about it but I think no probabilistic distribution is stated, so I think not) | No | No | Yes(Geographical, KNN, Probability) | No | No | No | Yes | No | Yes | No | No | Yes | Yes | No | No | No | No | No | Yes(1 randomly selected Check-in for each user. Leave one out) | No | No | No | No | Ranking | Precision and Recall | No(PCLR, PMM and USG) | No | No | Yes(at least USG) | No | Check-ins | No | Yes | No | No | No | No | Yes | Prev-No filtering | No | Gowalla: no further details | ||||||||||||||||||||||
52 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
53 | 2015 | ? | Point-of-interest recommendations using categorical information: An information retrieval perspective | Gao, Z., Huang, J., Zhou, E. | ??? | Journal of Computational Information Systems | Journal | ??? | 1 | 0 | IGNORE | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
54 | 2015 | ? | Local point of interest recommendation based on probability | Zhang, M., Yin, S.-Q., Sun, M.-M., Gao, S. | ??? | FSDM | Conference | 1 | 0 | IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
55 | 2015 | SSU | An efficient approach to generating location-sensitive recommendations in ad-hoc social network environments | Hao, F., Li, S., Min, G., Kim, H.-C., Yau, S.S., Yang, L.T. | https://ieeexplore.ieee.org/document/7036059 | IEEE Transactions on Services Computing | Journal | 1 | 48 | Not on the scope: location sensitive recommendation, not poi recommendation. IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
56 | 2015 | ---No-Acronym-- | A privacy-enhancing model for location-based personalized recommendations | Huang, J., Qi, J., Xu, Y., Chen, J. | https://link.springer.com/article/10.1007%2Fs10619-014-7148-8 | Distributed and Parallel Databases | Journal | 1,3 | 10 | Not on the scope: about privacy, not poi recommendation. IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
57 | 2015 | ---No-Acronym-- | Personalized trip recommendation with POI availability and uncertain traveling time | Zhang, C., Liang, H., Wang, K., Sun, J. | https://dl.acm.org/citation.cfm?id=2806558 | CIKM | Conference | 1,3 | 42 | Trip recommendation. IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
58 | 2015 | ---No-Acronym-- | Predicting the popularity of micro-reviews: A Foursquare case study | Marisa Vasconcelos, Jussara M. Almeida, Marcos André Gonçalves | https://www.sciencedirect.com/science/article/pii/S0020025515004843 | Information Sciences | Journal | 2 | 0 | Not POI recommendation: prediction of review popularity. IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
59 | 2015 | GA, GA+PO, GA+PO+CELF | On information coverage for location category based point-of-interest recommendation | Chen, X., Zeng, Y., Cong, G., Qin, S., Xiang, Y., Dai, Y. | http://www.aaai.org/ocs/index.php/AAAI/AAAI15/paper/view/9703 | AAAI | Conference | 1 | 30 | \cite{DBLP:conf/aaai/ChenZCQXD15} | It is not POI recommendation exactly. Main focus is on predicting category. IGNORE | Yes | No | No | No | No | No | No | No | No | Yes | No | No | No | No | Yes | Yes | No | No | No | No | No | Yes(random split, 80% train 20% test, for each user) | No | No | Yes(users and items with at least 5 POIs, or POIs that have at least 5 users) | Yes(Singapore for Foursquare and Austin for Gowalla) | Ranking | Precision, Recall and Diversity | No(GA, UST, Random) | Yes(Random) | No | Yes(UST) | No | POIs | No | Yes | Yes | No | No | No(They refere to other papers: Time-aware point-of-interest recommendation 2013, Friendship and mobility: user movement in location-based social networks 2011) | Yes | Post | No | Foursquare and Gowalla: Time-aware point-of-interest recommendation 2013 for Foursquare and Friendship and mobility: user movement in location-based social networks 2011 for Gowalla | |||||||||||||||||||||
60 | 2015 | LORE(repeated) | Spatiotemporal sequential influence modeling for location recommendations: A gravity-based approach | Zhang, J.-D., Chow, C.-Y. | https://dl.acm.org/citation.cfm?id=2786761 | TIST | Journal | 1 | 53 | \cite{DBLP:journals/tist/ZhangC15} | IGNORE. Repeated in 2014. Extension of paper \cite{DBLP:conf/gis/ZhangCL14} | No | No | Yes(Power-law) | No | Yes | Yes(Sequential, Temporal, Geographical, Mass model) | No | No | No | Yes | Yes | No | No | Yes | Yes | Yes | No | No | Yes(80% train, 20% test) | No | No | No | No | No | No | No | Ranking | Precision and Recall | No(STI, USG, CoRe, LCARS, DRW, FMC, AMC) | No | No | Yes(USG, CORE) | No | Check-ins | No | Yes | Yes | No | Brightkite | No(They refer to a couple of papers: Friendship and mobility: User movement in location- based social networks (2011) and gSCorr: Modeling geo-social correlations for new check-ins on location-based social networks (2012) | Yes | Prev-No filtering | No | Foursquare, Gowalla and Brighkite: they refer to Friendship and mobility: User movement in location- based social networks (2011) for Gowalla and Brighkite and gSCorr: Modeling geo-social correlations for new check-ins on location-based social networks (2012) for Foursquare | |||||||||||||||||||||
61 | 2015 | ---No-Acronym-- | A feasibility study of POI recommendation based on bursts of visits | Fukuda, T., Aritsugi, M. | https://dl.acm.org/citation.cfm?id=2837185.2837270 | iiWAS | Conference | 1,3 | 0 | \cite{DBLP:conf/iiwas/FukudaA15} | it is in the scope: they do POI recommendation using bursts | No | No | No | No | No | No | Yes(Burst detection) | No | No | No | No | No | No(we wont consider events as content) | No | No | Yes | Yes | No | No | Yes | No | No | No | No | No | No | No | Ranking | S (Success @ k) | No(Col-fil, loc-only, loc-time) | No | Yes(only loc is CF) | No | No | Check-ins | No | Yes | No | No | No | Yes(https://snap.stanford.edu/index.html. They also refer to Friendship and mobility: User movement in location-based social networks 2011) | Yes | Prev-No filtering | No | Gowalla: They refer to Friendship and mobility: User movement in location-based social networks 2011 and use this url https://snap.stanford.edu/index.html | ||||||||||||||||||||
62 | 2015 | EFC | Extended feature combination model for recommendations in location-based mobile services | Sattari, M., Toroslu, I.H., Karagoz, P., Symeonidis, P., Manolopoulos, Y. | https://link.springer.com/article/10.1007/s10115-014-0776-5 | Knowledge Information Systems | Journal | 1 | 11 | \cite{DBLP:journals/kais/SattariTKSM15} | Social activity recommendation | Yes | Yes | No | No | No | Yes | High Order Singular Value Decomposition (HOSVD) | No | No | No | No | No | No | No | No | Yes | No | No | No | No | No | No | Yes(10-fold cross validation in the dataset) | No | Yes(Gowalla dataset is filtered in order to have similar stats to Geosocial2data) | Yes(For Gowalla is in Paris and for Geosocial2 is in Greece) | Error/Ranking | MAE, RMSE, Precision, Recall, F1 | No(CB, CF, Hybrid) | No | Yes(CF, CB) | No | No(cross-validation) | Check-ins(not clear but the 10-fold is in the dataset) | No | Yes | No | No | Geosocial2 | Yes(http://snap.stanford.edu/data/loc-gowalla.html for Gowalla and http://delab.csd.auth.gr/geosocial2/index2.html for Geosocial) | Not complete. Number of POIs not stated | Prev-No filtering for Geosocial2 and Prev for Gowalla (we do not have the filtered stats for Gowalla) | No | Gowalla and GeoSocial2: http://snap.stanford.edu/data/loc-gowalla.html for Gowalla and http://delab.csd.auth.gr/geosocial2/index2.html for GeoSocial2 | |||||||||||||||||||||
63 | 2015 | gSCorr | Addressing the cold-start problem in location recommendation using geo-social correlations | Gao, H., Tang, J., Liu, H. | https://link.springer.com/article/10.1007/s10618-014-0343-4 | Data Mining and Knowledge Discovery | Journal | 1,3 | 35 | \cite{DBLP:journals/datamine/GaoTL15} | Yes | Yes | Yes | No | No | Yes(final sum of 4 probabilities. Some of the uses UBKnn) | Gradient method | No | No | Yes | Yes | No | No | No | Yes(I think yes. They incorporate temporal information there) | Yes | No | No | Yes(6 months train and seems to 1 month test. With different months) | No | No | No | No | No | No | No | Ranking | Accuracy | No(S.LF.UF, PSMM, SHM, CF) | No | Yes(CF) | Yes(S.LF.UF) | No | Check-ins | Yes(the paper is based on cold-start) | No | Yes | No | No | Yes(http://www.public.asu.edu/~hgao16/dataset.html) but this URL for Foursquare does not work | Yes | Post | No | Foursquare: http://www.public.asu.edu/~hgao16/dataset.html but url does not work | ||||||||||||||||||||||
64 | 2015 | URG+SM | Location recommendation incorporating temporal and spatial effects | Kojima, N., Takagi, T. | https://ieeexplore.ieee.org/document/7396816 | WI-IAT | Conference | 1,3 | 0 | \cite{DBLP:conf/webi/KojimaT15} | Yes | No | Yes(prob estimation based on distance, power law) | No | No | Yes(spatial and time influence) | No | No | No | Yes | No | No | No | No | Yes | Yes | No | No | No Information | No Information | No Information | No Information | No Information | No Information | No | Yes(Foursquare, Singapore and Gowalla California and Nevada) | Ranking | Precision and Recall | No(UTE+SE) | No | No | Yes(UTESE) | No | No information | No | Yes | Yes | No | No | No | Yes | Prev-No filtering | No | Foursquare and Gowalla: no further details | ||||||||||||||||||||||
65 | 2015 | TSLR + TSLRS | Topic-sensitive location recommendation with spatial awareness | Guo, Q., Huang, Y., Theng, Y.-L. | https://ieeexplore.ieee.org/document/7396810 | WI-IAT | Conference | 1,3 | 0 | \cite{DBLP:conf/webi/GuoHT15} | No | Yes(Topical pageRank and LDA) | Yes(Topical Page Rank and LDA) | No | Yes(Topical Page Rank) | No | No | No | No | Yes(for the second approach, TSLRS) | No | Yes | Yes | No | No | Yes | No | No | No | No | No | Yes(for each user, 80% of the Check-ins to train and rest to test) | No | No | Yes(users checked-in at least 10 locations) | Yes(Los Angeles, New York, San Francisco) | Ranking | Precision and Recall | No(UCF, PageRank, GCF, Topical PageRank) | No | Yes(UCF) | Yes(UCF+ geographical info) | No | Check-ins | No | Yes | No | No | No | Yes(http://snap.stanford.edu/index.html) | Yes | Post | No | Gowalla: They refer to Friendship and mobility: User movement in location-based social networks 2011 and use this url https://snap.stanford.edu/index.html | ||||||||||||||||||||||
66 | 2015 | RWCAR | Random walk based context-aware activity recommendation for location based social networks | Bagci, H., Karagoz, P. | https://ieeexplore.ieee.org/document/7344852 | DSAA | Conference | 1 | 13 | \cite{DBLP:conf/dsaa/BagciK15} | Activity recommendation | No | No | No | No | Yes | No | No | Not in the model, but for evaluation they use it (DBSCAN) | No | Yes(vecinity) | Yes | Yes, in the evaluation they use categories of Foursquare | No | No | No | Yes | No | No | No | Yes(weird evaluation methodology. Sort the timestamps and group them by clusters) | No | No | No | No | No | No | Ranking | Precision, Recall and F1 | No(PBAR, FBAR, EBAR | Yes(Pop) | No(friend-based) | No | No | Check-ins | No | No | Yes | No | No | No(They refer to this paper for Foursquare: gscorr: Modeling geo-social correlations for new check-ins on location-based social networks) | Yes | Prev-No filtering | No | Foursquare: they refer to gscorr: Modeling geo-social correlations for new check-ins on location-based social networks | |||||||||||||||||||||
67 | 2015 | FGLR | Point-of-interest recommendation in location-based social networks with personalized geo-social influence | Huang, L., Ma, Y., Liu, Y. | https://ieeexplore.ieee.org/document/7385525 | China Communications | Journal | ??? | 1 | 14 | \cite{Huang2015} | No | Yes(They update parameters and gradients) | Yes | No | No | No | Gradient descent | No | No | Yes | Yes | No | No | No | No | Yes | No | No | Yes(70% previous Check-ins to training, rest to test) | No | No | No | No | No | No | No | Ranking | Precision and Recall | No(NMF, MGM, SR, iGSLR, Core) | No | Yes(NMF) | Yes(MGM, igslr) | No | Check-ins | No | Yes | Yes | No | No | No(They refer to Friendship and mobility: user movement in location-based social networks 2011 for Gowalla and Exploring Social-Histori- cal Ties on Location-Based Social Networks for Foursquare) | Yes | Prev-No filtering | No | Foursquare and Gowalla: Friendship and mobility: user movement in location-based social networks 2011 for Gowalla and Exploring Social-Historical Ties on Location-Based Social Networks for Foursquare) | |||||||||||||||||||||
68 | 2015 | City Melange | Interactive multimodal learning for venue recommendation | Zahalka, J., Rudinac, S., Worring, M. | https://ieeexplore.ieee.org/document/7272105 | IEEE Transactions on Multimedia | Journal | 1 | 16 | \cite{DBLP:journals/tmm/ZahalkaRW15} | No | Yes(LDA - Topic Modeling) | Yes(LDA - Topic Modeling) | Yes(CNN) | No | Yes(SVM, topic modeling and Convolutional neural network. ) | Yes(SVMs...) | No | Yes | No | No | No | No | Yes | No | No | Yes | No(although they claim to use an online recommender and even in this case, interactive) | No | No | No | No | No | No | No | Yes(Weird random. For each city they select 100 random users and the 4-fold cross validation) | No | Yes(New York and Amsterdam) | Ranking | Precision and Recall and Time | Yes(PopRank, BPRMF, WRMF) | Yes(Pop) | Yes(BPRMF) | No | No | POIs | No | No | Yes | No | Yes(also use Flickr and Ficasa) | No | Not complete | Prev-No filtering | No | Fourquare and Flickr: no further details | ||||||||||||||||||||
69 | 2015 | REGULA | REGULA: Utilizing the regularity of human mobility for location recommendation | Mudda, S., Giordano, S. | https://dl.acm.org/doi/abs/10.1145/2833165.2833172?download=true | GeoStreaming/SIGSPATIAL | Conference | 1,3 | 3 | \cite{DBLP:conf/gis/MuddaG15} | No | No | No | No | No | Yes(Time, distance and friends) | No | No | No | Yes | Yes | No | No | No | Yes | Yes | No | No | Yes(different times in intervals for 30 days) | No | No | No | No | No | No | No | Ranking | Precision and Recall | No(LFBCA, LocCF, UserCF) | No | Yes(UB and LocCF (ib) | Yes(LFBCA) | No | Check-ins | No | Yes | No | No | Brightkite | No(They refer to Friendship and mobility: User movement in location-based social networks 2011) | Yes | Prev-No filtering | No | Gowalla and Brightkite: They refer to Friendship and mobility: User movement in location-based social networks 2011 | ||||||||||||||||||||||
70 | 2015 | LTSCR | Location and time aware social collaborative retrieval for new successive point-of-interest recommendation | Zhang, W., Wang, J. | https://dl.acm.org/citation.cfm?id=2806564 | CIKM | Conference | 1 | 55 | \cite{DBLP:conf/cikm/ZhangW15} | Next POI. Successive POI recommendation | No(They also use similarities but no neighbours) | Yes | No | No | No | No | Stochastic Gradient Descent | No | No | Yes(There are grids in the learning phase) | Yes | No | No | No(They discuss about transition patterns but it seems no) | Yes | Yes | No | No | No | No | Yes(70% training 10%validation and 20% test but it is not stated if it is at POI or Check-in level) | No | No | No | Yes(users with at least 10 records and POIs with at least 5 records) | No | Ranking | Precision and Recall | Yes(Pop, Pop+LR, PMF+LR, FPMC+LR) | Yes(Pop) | No(PMF with geographical) | Yes(Pop with geographical) | Yes | Check-ins | No | Yes | No | No | Brightkite | No(They refer to Friendship and mobility: User movement in location-based social networks 2011) | Yes | Post | No | Gowalla and Brightkite: They refer to Friendship and mobility: User movement in location-based social networks 2011 | |||||||||||||||||||||
71 | 2015 | GeoMF-TD | POI recommendation: Towards fused matrix factorization with geographical and temporal influences | Griesner, J.-B., Abdessalem, T., Naacke, H. | https://dl.acm.org/citation.cfm?id=2799679 | RecSys | Conference | 1 | 42 | \cite{DBLP:conf/recsys/GriesnerAN15} | No | Yes | No(I think not because although they claim to use gaussian distribution, they do not learn anything) | No | No | No | No | No | No | Yes | No | No | No | No | Yes | Yes | No | No | No | No | No | Yes(20%-40% of the locations to test for every user) | No | No | Yes(remove users with less than 50 Check-ins) | Yes(only Check-ins in France) | Ranking | Precision and Recall | No(GeoMF) | No | No | Yes(GeoMF) | No | POIs | No | Yes | No | No | No | Yes(http://snap.stanford.edu/data/loc-gowalla.html) | Yes | Prev and Post (i will indicate post) | No | Gowalla: http://snap.stanford.edu/data/loc-gowalla.html | ||||||||||||||||||||||
72 | 2015 | USPB | Adaptive location recommendation algorithm based on location-based social networks | Lin, K., Wang, J., Zhang, Z., Chen, Y., Xu, Z. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7250231 | ICCSE | Conference | ??? | 1 | 13 | \cite{Lin2015} | Yes | No | Yes | No | No | Yes(very similar to the USG, bayesian, social and KNN) | No | No | No | Yes(distance in the naive bayes) | Yes(friends) | No | No | No | No | Yes | No | No | No | No | No | Yes(weird evaluation. Select a random location and repeat the same procedure for every user. Leave one out ) | No | No | No | No | Ranking | Precision | No(U, S, B, USG) | No | Yes(U) | Yes(USG) | No | Check-ins | Yes | No | Yes | No | No | No(They refer to An Empirical Study of Geographic User Activity Patterns in Foursquare 2011) | Not complete. We only know the number of Check-ins | Prev-No filtering | No | Foursquare: They refer to An Empirical Study of Geographic User Activity Patterns in Foursquare 2011 | |||||||||||||||||||||
73 | 2015 | iGeoRec | IGeoRec: A Personalized and Efficient Geographical Location Recommendation Framework | Zhang, J.-D., Chow, C.-Y., Li, Y. | https://ieeexplore.ieee.org/document/6824843 | IEEE Transactions on Services Computing | Journal | 1 | 59 | \cite{DBLP:journals/tsc/ZhangCL15} | No | No | Yes | No | No | No | No | Yes | No | Yes | Yes | No | No | No | No | Yes | No | No | Yes(half of the data for training with different percentages and rest for test) | No | No | No | No | No | No | No | Ranking | Precision and Recall | No(NMF, MGM, PD) | No | Yes(NMF is matrix factorization) | Yes(MGM) | No | Check-ins | Yes | Yes | Yes | No | No | Yes(http://www.public.asu.edu/~hgao16/Publications.html for Foursquare, although it does not work and http:// snap.stanford.edu/data/loc-gowalla.html) | Yes | Prev-No filtering | No | Foursquare and Gowalla: They refer to gSCorr: Modeling geo-social correlations for new check-ins on location-based social networks 2011 and Friendship and mobility: User movement in location-based social networks 2011 for Gowalla, for Foursquare: http://www.public.asu.edu/~hgao16/Publications.html and for Gowalla http://snap.stanford.edu/data/loc-gowalla.html) | ||||||||||||||||||||||
74 | 2015 | TenInt | Context-aware point-of-interest recommendation using Tensor Factorization with social regularization | Yao, L., Sheng, Q.Z., Qin, Y., Wang, X., Shemshadi, A., He, Q. | https://dl.acm.org/citation.cfm?id=2767794 | SIGIR | Conference | 1 | 69 | \cite{DBLP:conf/sigir/YaoSQWSH15} | No | Yes | No | No | No | No | CANDECOMP/PARAFAC (CP) descomposition | No | No | No | Yes | No | No | No | Yes | Yes | No | No | No | No | No | Yes(20% of the POIs to test for every user) | No | No | Yes(remove users with less than 5 POIs visited and removed POIs with less than 5 different users) | No | Ranking | Precision and Recall | No(NMF, UCF,ICF,FA, GA, Time-aware) | No | Yes(UBCF, IBCF) | Yes(Geographical aware) | No | POIs | No | No | No | No | Brightkite | Yes(http://snap.stanford.edu/data/loc-brightkite.html and They also refer to this paper . Friendship and mobility: user movement in location-based social networks. 2011) | No(as they refer to the original Brightkite dataset, we can take the stats from that dataset?) | None | No | Brighkite: http://snap.stanford.edu/data/loc-brightkite.html and They also refer to this paper . Friendship and mobility: user movement in location-based social networks. 2011 | ||||||||||||||||||||||
75 | 2015 | TCL-K | Learning recency based comparative choice towards point-of-interest recommendation | Li, X., Xu, G., Chen, E., Zong, Y. | https://www.sciencedirect.com/science/article/abs/pii/S095741741500069X | Expert Systems with Applications | Journal | 1,2,3 | 21 | \cite{DBLP:journals/eswa/LiXCZ15} | No | Yes | Yes(seems a probabilistic MF) | No | No | No | Bayesian optimization | No | No | No | No | No | No | No | Yes | Yes | No | No | No | Yes(time window for every user) | No | No | No | No | Yes(users with less than 10 ratings for Yelp and 8 for tripadvisor) | No | Ranking | Recall and MAP | No(TBCF, SPLINE, PMF, BPR) | No | Yes(BPR) | No | No | POIs | No | No | No | Yes | Yes(Tripadvisor) | Yes(https://www.yelp.com.au/dataset/challenge for yelp and http://sifaka.cs.uiuc.edu/wang296/Data/index.html for tripadvisor, although the url does not work) | Yes | Post | Yes(https://www.dropbox.com/sh/f8usnw19yg7werv/AABlR7ipr4uLbrtUHmlMkDwja?dl=0) | Yes | No | Yes(for their approach) | Yelp and TripAdvisor: https://www.yelp.com.au/dataset/challenge for Yelp and http://sifaka.cs.uiuc.edu/wang296/Data/index.html for TripAdvisor | |||||||||||||||||||
76 | 2015 | LA-LDA | Modeling location-based user rating profiles for personalized recommendation | Yin, H., Cui, B., Chen, L., Hu, Z., Zhang, C. | https://dl.acm.org/citation.cfm?id=2663356 | TKDD | Journal | 1,3 | 105 | \cite{DBLP:journals/tkdd/YinCCHZ15} | It not only POI recommendation | No | Yes(LDA) | Yes(LDA) | No | No | No | No | No | No | Yes | No | Yes | No | No | No | Yes | No(although they claim to use an online recommender) | No | No | No | No | Yes(for every user, 70% training, 10% validation and 30% test for every location) | No | No | No | No(Even if Doubant event is for cities in China, the datasets are no restricted) | Ranking | Hit@k and EDE (expected distance Error) | No(LARS, Geographical probabilistic factor model, LDA and other versions of the algorithm) | No | No | Yes(Geographical factor model) | Yes | POIs | Yes | Yes | No | No | Yes(Movielens, DoubanEvent) | No(They refer to this paper for Gowalla: . Friendship and mobility: User movement in location- based social networks 2011 and this url https://www.datatang.com/Member/67341 that does not work) | Not complete. Number of POIs not stated | Prev-No filtering | No | Gowalla, DoubanEvent and Movielens: Gowalla, Friendship and mobility: User movement in location- based social networks 2011 DoubanEvent no information | |||||||||||||||||||||
77 | 2015 | LURA | Personalized location recommendation by aggregating multiple recommenders in diversity | Lu, Z., Wang, H., Mamoulis, N., Tu, W., Cheung, D.W. | http://ceur-ws.org/Vol-1405/paper-05.pdf | LocalRec/RecSys | Conference | 1 | 2 | \cite{DBLP:conf/recsys/LuWMTC15} | Yes(part of the hybrid) | Yes(Implicit matrix factorization) | Yes(part of the hybrid) | No | No | Yes(combines different location recommenders) | Gradient descent | No | No | Yes | Yes | Yes | No | No | Yes(Time weighted CF) | Yes | No | No | Yes([1, t -deltat], for training, (t-deltat, t] for validation and (t, t+ deltat) for test | No | No | No | No | No | No | No | Ranking | Precision and Recall | No(UCF, FCF, FLCF,GCF,CCF,ICF,TCF,PLM,KDM,SKDM,IMF, USG, iGSLR,BPRMF,GeoMF,SBPR,RankBoost-SA, RankBoost-RA) | No | Yes(UCF, BPR) | Yes(IGSLR, GeoMF) | Yes | Check-ins | No | Yes | Yes | No | No | No | Yes | Post | No | Gowalla and Foursquare: o info | ||||||||||||||||||||||
78 | 2015 | UGC | Semantic-based location recommendation with multimodal venue semantics | Wang, X., Zhao, Y.-L., Nie, L., Gao, Y., Nie, W., Zha, Z.-J., Chua, T.-S. | https://ieeexplore.ieee.org/document/6996042 | IEEE Transactions on Multimedia | Journal | 1 | 55 | \cite{DBLP:journals/tmm/WangZNGNZC15} | Yes(They say they use neighbours) | Yes | No | No | No | No | Gradient Descent | No | No | No | No | Yes | Yes | No | No | Yes | No | No | No | No | No | Yes(for each user 50% Check-ins to training rest to test) | No | No | Yes(they select a subset) | Yes(Singapore) | Ranking | NDCG, F1, Precision, Recall | No(NMF and PMF) | No | Yes(NMF) | No | No | Check-ins | No | No | Yes | No | No | No | Yes | Post | No | Foursquare: no info provided | ||||||||||||||||||||||
79 | 2015 | HRWR | Deriving an effective hypergraph model for point of interest recommendation | Qi, M., Li, X., Liao, L., Song, D., Cheung, W.K. | https://link.springer.com/chapter/10.1007/978-3-319-25159-2_71 | KSEM | Conference | 1,3 | 0 | \cite{DBLP:conf/ksem/QiLLSC15} | No | No | No(If we consider transition probability. I think it is not) | No | Yes(Random Walk with restart and They also use PageRank) | No | No | No | No | No | Yes | Yes | No | No(They say something about transition probability but I think we should not consider it in this case) | No | Yes | No | No | No information provided | No information provided | No information provided | No information provided | No information provided | No information provided | No | No | Ranking | Precision and Recall | No(CF, Supervised Random Random Walk, RWR) | No | Yes(CF) | No | No | No Information | No | No | Yes | No | No | No(but they refer to this paper Location-based and preference-aware recom- mendation using sparse geo-social networking data 2012 | No(As they refer to other paper, we can extract the statistics from the other paper) | None | No | Foursquare: They refer to Location-based and preference-aware recommendation using sparse geosocial networking data 2012 | ||||||||||||||||||||||
80 | 2015 | ---No-Acronym-- | Complementary usage of tips and reviews for location recommendation in Yelp | Gupta, S., Pathak, S., Mitra, B. | https://link.springer.com/chapter/10.1007/978-3-319-18032-8_56 | PAKDD | Conference | 1 | 5 | \cite{DBLP:conf/pakdd/GuptaPM15} | No | No | No | No | Yes | No | No | No | No | Yes | Yes | No | Yes(textual) | No | No | Yes | No | No | No | No | No | Yes(for every user 20%-40% of the locations are sent to test) | No | No | Yes(keep users with at least 1 tip or review for at least 30 locations) | Yes(Phoenix) | Ranking | Precision and Recall | No(CF, Graph-based, NMF) | No | Yes(CF) | No | No | POIs | No | No | No | Yes | Brightkite | Yes(https://www.yelp.com/dataset/challenge) | Not complete (not Check-ins, just reviews and no information for Brighkite) | Prev | No | Yelp and Brightkite: for yelp, they refer to https://www.yelp.com/dataset/challenge for Brighkite they refer to Friendship and mobility: user movement in location-based social networks 2011 | ||||||||||||||||||||||
81 | 2015 | MARS | MARS: A multi-aspect Recommender system for Point-of-Interest | Li, X., Xu, G., Chen, E., Li, L. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7113395 | ICDE | Conference | 1 | 9 | \cite{DBLP:conf/icde/LiXCL15} | Case study for evaluation | No | Yes | No | No | No | No(linear combination of individual and collaborative utility). | Gradient descent | No | No | No | No | No | Yes(review) | No | No | No | No | Yes | No | No | No | No | No | No | Yes(User online study) | Yes(removed user and POIs with less than 5 reviews) | Yes(Phoenix) | None | None | No | No | No | No | No | No Information | No | No | No | Yes | No | Yes(http://www.yelp.com.au/dataset challenge) | Yes | Post | No | Yelp: http://www.yelp.com.au/dataset challenge | ||||||||||||||||||||
82 | 2015 | TA-FPMC | Crafting a time-aware point-of-interest recommendation via pairwise interaction tensor factorization | Zhao, X., Li, X., Liao, L., Song, D., Cheung, W.K. | https://link.springer.com/chapter/10.1007/978-3-319-25159-2_41 | KSEM | Conference | 1 | 6 | \cite{DBLP:conf/ksem/ZhaoLLSC15} | Category recommendation AND POI recommendation | No | Yes | Yes | No | No | No | Gradient descent | No | No | Yes | No | Yes | No | Yes | Yes | Yes | No | No | No | No | Yes(80% training 20% test of Check-ins. However it is not specifically stated that it is random) | No | No | No | Yes(filter out inactive users) | Yes(Los Angeles) | Ranking | Precision and Category Precision | No(FPMC-LR, FPMC, Random, MF) | Yes(Random) | Yes(MF) | No | No | Check-ins | No(they discuss about it but not specific experiment is conducted) | No | Yes | No | No | No (They refer to this paper: Location-based and preference-aware recom- mendation using sparse geo-social networking data 2012) | Yes | Prev | No | Foursquare: They refer to Location-based and preference-aware recom- mendation using sparse geo-social networking data 2012 | |||||||||||||||||||||
83 | 2015 | CoRe | CoRe: Exploiting the personalized influence of two-dimensional geographic coordinates for location recommendations | Zhang, J.-D., Chow, C.-Y. | https://www.sciencedirect.com/science/article/pii/S0020025514009207 | IS | Journal | 1,2 | 77 | \cite{DBLP:journals/isci/ZhangC15} | Yes | No | Yes | No | No | Yes(Fusing KDE and social) | No | No | No | Yes | Yes | No | No | No | No | Yes | No | No | Yes(50% Check-ins to train rest to test) | No | No | No | No | No | No | No | Ranking | Precision and Recall | No(MGM, PD, iGSLR, Exact, SCF, Prod, Sum) | No | No | Yes(MGM, igslr, Power-law) | No | Check-ins | No | Yes | Yes | No | No | Yes(http://www.public.asu.edu/~hgao16/Publications.html for FOursqaue, although it does not work and http:// snap.stanford.edu/data/loc-gowalla.html) | Yes | Prev-No filtering | No | Foursquare and Gowalla: Friendship and mobility: user movement in location-based social networks 2011 for Gowalla and gSCorr: Modeling geo-social correlations for new check-ins on location-based social networks 2012 for Foursquare) | ||||||||||||||||||||||
84 | 2015 | STS Location Recommender | Unifying spatial, temporal and semantic features for an effective GPS trajectory-based location recommendation | Abdel-Fatao, H., Li, J., Liu, J. | https://link.springer.com/chapter/10.1007/978-3-319-19548-3_4 | ADC | Conference | 1 | 5 | \cite{DBLP:conf/adc/Abdel-FataoLL15} | Yes | No | No | No | No | Yes(Preference score Estimation + User CF) | No | Yes(OPTICS) | No | Yes | No | Yes | No | No | Yes | Yes | No | No | No | No | No | Yes(random 30% -40% and 50% of the locations for test for each user) | No | No | Yes(from 182 to 149 users) | No | Error/Ranking | Precision Recall and RMSE | No(UBCF and UCLAF) | No | Yes(UBCF) | No | No | POIs | No | No | No | No | GeoLife | No(They refer to this url but i cannot find the dataset here http://research.microsoft.com/en-us/downloads/b16d359d-d164-469e-9fd4-daa38f2 b2e13/ | Not complete (only users) | None | No | GeoLife: https://www.microsoft.com/en-us/research/products/?from=http%3A%2F%2Fresearch.microsoft.com%2Fen-us%2Fdownloads%2Fb16d359d-d164-469e-9fd4-daa38f2 | ||||||||||||||||||||||
85 | 2015 | Orec | ORec: An opinion-based point-of-interest recommendation framework | Zhang, J.-D., Chow, C.-Y., Zheng, Y. | https://dl.acm.org/doi/10.1145/2806416.2806516 | CIKM | Conference | 1 | 42 | \cite{DBLP:conf/cikm/ZhangCZ15} | Yes | No | Yes | No | No | Yes(Geographical, Social, KNN) | No | Yes(aspect cluster) | No | Yes | Yes | No(although they use aspects, they are extracted from text) | Yes(tips) | No | No | Yes | No | No | No | No | No | No | Yes(Although it not very clear. We will assume cross-validation CC) | No | No | Yes(for Foursquare, for yelp no) | Ranking | Precision and Recall | No(IrenMF, PD, iGSLR, LCARS, UAI, ORec) | No | No | Yes(IRENMF, IGSLR) | No(Cross-validation) | POIs(Equivalent) | No | No | Yes | Yes | No | Yes(http://www.yelp.com/dataset_challenge, from 2014 and refering to paper Location-based and preference-aware recommendation using sparse geo-social networking data 2012) | No | None | No | Yelp and Foursquare: http://www.yelp.com/dataset_challenge from 2014 and Location-based and preference-aware recommendation using sparse geo-social networking for Foursquare 2012 | ||||||||||||||||||||||
86 | 2015 | Poisson Geo-PFM | A general geographical probabilistic factor model for point of interest recommendation | Liu, B., Xiong, H., Papadimitriou, S., Fu, Y., Yao, Z. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=6919280 | TKDE | Journal | 1 | 112 | \cite{DBLP:journals/tkde/LiuXPFY15} | Yes | No | Yes(Poisson MF) | Yes(Poisson MF) | No | No | No | Gradient descent, Expectation Maximization | Yes | No | Yes | No | No | No | No | No | Yes | No | No | No | No | Yes(80% training 20% test of Check-ins) | No | No | No | Yes(For gowalla) | Yes(Foursquare uses USA, but for the other, no info) | Ranking | Precision and Recall | No (PMF, BNMF, Poisson Factor Model, Fu-POIMF) | No | Yes(PMF) | Yes(Geo-NMF) | No | Check-ins | No | Yes | Yes | No | Brightkite | No(They refer to other papers “Friendship andmobility: User movement in location-based social networks 2011, Exploring millions of footprints in location sharing services 2011 and Toward traffic- driven location-based web search 2011) | Yes | Prev-No filtering for some and Post for Gowalla | No | Foursquare and Gowalla and Brightkite: They refer to Friendship and mobility: User movement in location-based social networks 2011 for Gowalla and Brighttkite and Exploring millions of footprints in location sharing services 2011 and Toward traffic- driven location-based web search 2011 for Foursquare | |||||||||||||||||||||
87 | 2015 | RankGeoFM | Rank-geoFM: A ranking based geographical factorization method for point of interest recommendation | Li, X., Cong, G., Li, X.-L., Pham, T.-A.N., Krishnaswamy, S. | https://dl.acm.org/doi/10.1145/2766462.2767722 | SIGIR | Conference | 1 | 220 | \cite{DBLP:conf/sigir/LiCLPK15} | If it uses the temporal aspect. Can be considered as Time Aware | Yes | No | Yes | No | No | No | No | Gradient Descent, Ordered Weighted Pairwise Classification (OWPC) | No | No | Yes | No | No | No | No | Yes | Yes | No | No | No | Yes(70% for training, 10% for validation and 20% of Check-ins to test FOR EVERY USER) | No | No | No | No | No | Yes(Nevada and California for Gowalla, Singapore for Foursquare) | Ranking | Precision and Recall | No(UCF, UCF+G,PMF, BPRMF, GTBNMF, GeoMF, UCF, UTF, UCLAF, PITF, LRT, UTE+SE, BPP) | No | Yes(CF) | Yes(GeoMF, UCF-G) | Yes | Check-ins | No | Yes | Yes | No | No | Yes(http://www.ntu.edu.sg/home/gaocong/data /poidata.zip) Url does not work | Yes | Prev-No filtering | No | Foursquare and Gowalla: http://www.ntu.edu.sg/home/gaocong/data /poidata.zip for both and refer to Time-aware point-of-interest recommendation 2013 for also both | ||||||||||||||||||||
88 | 2015 | GeoSoca | GeoSoCa: Exploiting geographical, social and categorical correlations for point-of-interest recommendations | Zhang, J.-D., Chow, C.-Y. | https://dl.acm.org/doi/10.1145/2766462.2767711 | SIGIR | Conference | 1,3 | 196 | \cite{DBLP:conf/sigir/ZhangC15} | Yes | No | No | Yes(KDE and power-law) | No | No | Yes(Social, Categorical and Geographical) | No | No | No | Yes | Yes | Yes | No | No | No | Yes | No | No | Yes(50% train and 50% test by Check-ins) | No | No | No | No | No | No | Yes(Foursquare worldwide and Yelp Arizona) | Ranking | Precision and Recall | No(USG, CoRe, DRW, LCARS, NPCD) | No | No | Yes(USG) | No | Check-ins | No | No | Yes | Yes | No | Yes(they redirect to other paper and this url http://www.yelp.com/dataset_challenge in 2014) | No(shall we use the same stats as the paper from they are extracted?) | Prev-No filtering | No | Foursquare and Yelp: http://www.yelp.com/dataset_challenge for Yelp and refer to Content-aware point of interest recommendation on location-based social networks 2015 for Foursquare | |||||||||||||||||||||
89 | 2015 | PRME-G | Personalized ranking metric embedding for next new POI recommendation | Feng, S., Li, X., Zeng, Y., Cong, G., Chee, Y.M., Yuan, Q. | https://dl.acm.org/doi/10.5555/2832415.2832536 | IJCAI | Conference | 1,3 | 217 | \cite{DBLP:conf/ijcai/FengLZCCY15} | Next POI recommendation | Yes | No | Yes(I think yes, they use embeddings and latent space) | Yes | No | No | No(Although they claim to combine 2 things, the two are in the same model (optimize in BPR) | BPR, gradient descent | No | Yes(Metric embedding) | Yes | No | No | No | Yes | Yes(I think yes, they compute diferences between timestamps) | Yes | No | No | Yes(10 months train, 1 validation rest test) | No | No | No | No | No | Yes(removed users and items with less than 10 locations visites/users that have visited) | Yes(Singapore, Nevada and California) | Ranking | Precision and Recall | Yes(Pop, UCF, MF, MC, PME, FPMC) | Yes(Pop) | Yes(CF) | No | Yes | Check-ins | No | Yes | Yes | No | No | No(They refer to other papers: Time-aware point-of-interest recommendation 2013 and Friendship and mobility: user movement in location-based social networks (2011) | Yes | Post | No | Foursquare and Gowalla: they refer to Time-aware point-of-interest recommendation 2013 for Foursquare and Friendship and mobility: user movement in location-based social networks (2011) for Gowalla | ||||||||||||||||||||
90 | 2015 | CAPRF | Content-aware point of interest recommendation on location-based social networks | Gao, H., Tang, J., Hu, X., Liu, H. | https://www.aaai.org/ocs/index.php/AAAI/AAAI15/paper/view/9560/9456 | AAAI | Conference | 1 | 203 | \cite{DBLP:conf/aaai/GaoTHL15} | Yes | No | Yes | No | No | No | No | Gradient descent | No | No | No | No | No | Yes(it use tips/text. It fits this category) | No | No | Yes | No | No | No | No | No | Yes(Random 20% POIs to test for every user) | No | No | Yes(users with at least 2 different POIS) | Yes(California) | Ranking | Precision and Recall | No(UCF, PMF, NMF, STLR, SELR) | No | Yes(MF) | No | No | POIs | No | No | Yes | No | No | No | No | None | No | Foursquare: they say they follow the same procedure of some papers but nothin more | |||||||||||||||||||||
91 | 2015 | SFPMF and UIPMF | Point-of-interest recommender systems: A separate-space perspective | Li, H., Hong, R., Zhu, S., Ge, Y. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7373327 | ICDM | Conference | 1,3 | 26 | \cite{DBLP:conf/icdm/LiHZG15} | No | Yes | Yes | No | No | Yes(Equation 11 I think is the combination of both models) | Alternating Least Squares | No | No | Yes | Yes | No | No | No | No | Yes | No | No | No | Yes(80% of aggregted Check-ins of the user to train, rest to test) | No | No | No | No | Yes(Users visiting less than 5 removed and POIs with less than 5 users removed. Also removed users visiting more than 1000 locations) | No | Ranking | Precision, Recall and MAP | No(UB-KNN, USG, LOCABAL, RegPMF, PMF) | No | Yes(UC) | Yes(USG) | No | POIs (REMOVING REPETITIONS) | No | Yes | No | No | No | No(Refer to paper: Exploiting place features in link prediction on location-based social networks 2011) | Yes | Post | No | Gowalla: They refer to “Exploiting place features in link prediction on location-based social networks 2011 | ||||||||||||||||||||||
92 | 2015 | ICCF | Content-aware collaborative filtering for location recommendation based on human mobility data | Lian, D., Ge, Y., Zhang, F., Yuan, N.J., Xie, X., Zhou, T., Rui, Y. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7373330 | ICDM | Conference | 1,3 | 60 | \cite{DBLP:conf/icdm/LianGZYXZR15} | No | Yes | No | No | No | No | Alternating Least Squares | No | No | No | No | Yes(semantic) | Yes(text content) | No | No | Yes | No | No | No | No | No | No | 5-fold cross-validation Fix or CC. ONLY STATE NFOLD, ANYHTING ELSE | 5-fold cross-validation Fix or CC. ONLY STATE NFOLD, ANYHTING ELSE | Yes(users and items with at least 10 POIs, or POIs that have at least 10 users) | No | Ranking | Precision and Recall | No(LibFM, ICF, BPRMF, MMMF, PFM, GaP) | No | Yes(BPRMF) | No | No(Cross-valdiation) | Check-ins | Yes(I think they refer to new users) | No | No | No | Jiepang (chinese LBSN similar to Foursquare) | No | Not complete (number of Check-ins not stated) | Post | No | Jiepang: not much information | ||||||||||||||||||||||
93 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
94 | 2016 | ? | Textual-geographical-social aware point-of-interest recommendation | Xingyi, R., Meina, S., Haihong, E., Junde, S. | https://www.sciencedirect.com/science/article/abs/pii/S1005888516600663 | Journal of China Universities of Posts and Telecommunications | Journal | ??? | 1,2 | 4 | IGNORE | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
95 | 2016 | ? | GPS-based personalized point-of-interest recommendation algorithm | Zhang, Z., Pan, H. | ??? | Conference | ??? | 1 | 0 | IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
96 | 2016 | ? | Point-of-interest recommendation algorithm based on user similarity in location-based social networks | Tang, N., Lin, J., Weng, W., Zhu, S. | ??? | ICEB | Conference | ??? | 1 | 0 | IGNORE | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
97 | 2016 | ? | Time-aware collaborative location recommendation in location-based social networks | Zhang, M., Yin, S., Gao, S., Han, Z. | ??? | ICIC Express Letters | Journal | ??? | 1 | 0 | IGNORE | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
98 | 2016 | ? | A community-based hybrid location recommendation system in location-based social networks | Madhu, K.P., Manjula, D. | ??? | Asian Journal of Information Technology | Journal | ??? | 1 | 0 | IGNORE | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
99 | 2016 | ? | Personalized sequential point of interest recommendation on big social media | Sabu, M.M., Santhanakrishnan, C. | ??? | Journal | ??? | 1 | 0 | IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
100 | 2016 | TRP | Personalized Travel Package with Multi-Point-of-Interest Recommendation Based on Crowdsourced User Footprints | Yu, Z., Xu, H., Yang, Z., Guo, B. | https://ieeexplore.ieee.org/document/7145457 | IEEE Transactions on Human-Machine Systems | Journal | 1 | 153 | Trip recommendation. IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
101 | 2016 | ---No-Acronym-- | A study and analysis of recommendation systems for location-based social network (LBSN) with big data | Narayanan, M., Cherukuri, A.K. | https://www.sciencedirect.com/science/article/pii/S0970389616000021 | IIMB Management Review | Journal | ??? | 1,2 | 22 | I think it is not on the scope. IGNORE. It is like a mini-survey | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
102 | 2016 | ---No-Acronym-- | Friend recommendation algorithm based on location-based social networks | Lin, K., Chen, Y., Li, X., Wu, Q., Xu, Z. | https://ieeexplore.ieee.org/document/7883056 | ICSESS | Conference | ??? | 1 | 7 | Friend recommendation, not POI. IGNORE | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
103 | 2016 | ---No-Acronym-- | Trip Recommendation Meets Real-World Constraints: POI Availability, Diversity, and Traveling Time Uncertainty | Zhang, C., Liang, H., Wang, K. | https://dl.acm.org/citation.cfm?id=2986034.2948065 | TOIS | Journal | 1,3 | 27 | Trip recommendation. IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
104 | 2016 | ---No-Acronym-- | Who wants to join me? Companion recommendation in location based social networks | Liao, Y., Lam, W., Jameel, S., Schockaert, S., Xie, X. | https://dl.acm.org/citation.cfm?id=2970420 | ICTIR | Conference | 1,3 | 8 | It is companion recommendation aka friend recommendation. IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
105 | 2016 | ---No-Acronym-- | Friend Recommendation Algorithm for Online Social Networks Based on Location Preference | Wu, M., Wang, Z., Sun, H., Hu, H. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7726187 | ICISCE | Conference | ??? | 1 | 11 | Friend recommendation, not POI. IGNORE | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
106 | 2016 | PCGR | Personalized Group Recommender Systems for Location- and Event-Based Social Networks | Purushotham, S., Jay Kuo, C.-C. | https://dl.acm.org/citation.cfm?id=2987381 | TSAS | Journal | 1,3 | 18 | It is group recommendation. Maybe it is not on the scope. IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
107 | 2016 | RWCFR | Context-Aware Friend Recommendation for Location Based Social Networks using Random Walk | Hakan Bagci and Pinar Karagoz | https://dl.acm.org/doi/abs/10.1145/2872518.2890466 | WWW | Conference | 3 | 0 | Friend recommendation, not POI. IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
108 | 2016 | ---No-Acronym-- | Participatory Cultural Mapping Based on Collective Behavior Data in Location-Based Social Networks | Dingqi Yang, Daqing Zhang, Bingqing Qu | https://dl.acm.org/doi/10.1145/2814575 | TIST | Journal | 3 | 0 | Not recommending POIs. No POI recommendation method proposed. IGNORE --> analysis | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
109 | 2016 | ---No-Acronym-- | Providing recommendations on location-based social networks | Kosmides, P., Demestichas, K., Adamopoulou, E., Remoundou, C., Loumiotis, I., Theologou, M., Anagnostou, M. | https://link.springer.com/article/10.1007/s12652-016-0346-7 | Journal Ambient Intelligence and Humanized Computing | Journal | 1 | 15 | \cite{DBLP:journals/jaihc/KosmidesDARLTA16} | IGNORE. Repeated in 2014. Same as in \cite{Kosmides2014} | No | No | Yes(it is a probabilistic neural network) | Yes | No | Yes(they also use a k-means for clustering) | No(SVM, RBF as other approaches, but the model is the PNN) | Expectation Maximization | Yes | No | Yes | Yes | No | No | No | Yes | Yes | No | No | No | No | No | No | Yes(10-cross validation somewhere. I will state nFold CC) | No | No | No | None | Misclassification percentage | No(USG) | No | No | Yes(USG) | No | No information | No | No | Yes | No | No | No(They refer to this paper: Lars: a location-aware recommender system 2012) | No | None | No | Foursquare: They refer to Lars: a location-aware recommender system 2012 | ||||||||||||||||||||
110 | 2016 | CLoRW | Context-aware location recommendation by using a random walk-based approach | Bagci, H., Karagoz, P. | https://link.springer.com/article/10.1007/s10115-015-0857-0 | Knowledge Information Systems | Journal | 1,3 | 31 | \cite{DBLP:journals/kais/BagciK16} | IGNORE. Repeated in 2015. Same as in \cite{DBLP:conf/dsaa/BagciK15} | No | No | No(They say something bout transition probabilities but they are not shown) | No | Yes | No | No | Yes(DBSCAN) | No | Yes | Yes | No | No | No | No | Yes | No | No | No | No | Yes | No | No | No | No | No | Ranking | Precision, Recall and F1 Measure | Yes(Pop, Friend, Expert, CF, US) | Yes(popular) | Yes(CF) | Yes(USG) | No | Check-ins | No | Yes | Yes | No | Brightkite | No(They refer to paper: Friendship and mobility: user movement in location-based social networks 2011 and Gscorr: modeling geo-social correlations for new check-ins on location- based social networks 2012 | Yes | Post | No | Brightkite, Gowalla and Foursquare: They refer to Friendship and mobility: user movement in location-based social networks 2011 for Gowalla and Brighkite and Gscorr: modeling geo-social correlations for new check-ins on location- based social networks 2012 for Foursquare | |||||||||||||||||||||
111 | 2016 | sPCLR-DTW, sPCLR-BCC | Similarity-based probabilistic category-based location recommendation utilizing temporal and geographical influence | Zhou, D., Rahimi, S.M., Wang, X. | https://link.springer.com/article/10.1007/s41060-016-0011-4 | International Journal of Data Science and Analytics volume | Journal | 1 | 6 | IGNORE. Repeated in 2014. Same as \cite{DBLP:conf/dsaa/ZhouW14} | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
112 | 2016 | ---No-Acronym-- | A contextualized and personalized model to predict user interest using location-based social networks | Ming Li, Günther Sagl, Lucy Mburu, Hongchao Fan | https://www.sciencedirect.com/science/article/pii/S0198971516300357 | Computers, Environment and Urban Systems | Journal | 2 | 0 | \cite{DBLP:journals/urban/LiSMF16} | Not poi recommendation. It models user interest. --> it presents an algorithm, it should be considered (althout the metrics are not standard) | No | No | No(They discuss about probabilities but i do not see anything more. For me it is not) | No | No | No | Yes(Multinomial regression) | No | No | No | Yes | No | Yes | No | Yes | Yes | Yes | No | No | Yes | No | No | No | No | No | No | Yes(Chicago, Los Angeles, New York) | Error | CCR, Deviance, AIC, McFadden, CoxSnell | No(Only parts of the approach) | No | No | No | No | Check-ins | No | No | Yes | No | No | No | Not complete | None | No | Foursquare: no info provided | ||||||||||||||||||||
113 | 2016 | ---No-Acronym-- | A motivation-aware approach for point of interest recommendations | Vakeel, K.A., Ray, S. | http://ceur-ws.org/Vol-1685/paper4.pdf | RecTour/RecSys | Conference | 1 | 0 | \cite{DBLP:conf/recsys/VakeelR16} | I think it is not on the scope as it does not propose any approach, only argue to take into account motivation. --> there is an algorithm, although the evaluation is not offline. Considering in the survey | No | No | No | No | No | No | Yes(I marked as other. It is a post-classical recommendation approach. Refines the recommendatins from a classic recommender) | No | No | No | No(I think it is only on the recommendation) | No(It is part of the 7 motivations proposed but i do not see how it is modeled. Although the motivation is social, it is not pure social) | Yes(categorical of the users) | No | No | No | No | No | Yes | No | No | No | No | No | No | Yes(Case-study (10 users)) | No | No | None | No | No | No | No | No | No | None | No | No | Yes | No | No | No | No | None | No | Foursquare: only 10 users matching their criteria | |||||||||||||||||||
114 | 2016 | ---No-Acronym-- | Quality models for venue recommendation in location-based social network | Nie, W., Liu, A., Zhu, X., Su, Y. | https://link.springer.com/article/10.1007/s11042-014-2339-x | Multimedia Tools Applicantion | Journal | 1 | 2 | \cite{DBLP:journals/mta/NieLZS16} | I think it is not on the scope. --> there is an algorithm, although the evaluation is not offline | No | No | No | No | Yes(bipartite matching) | No | No | Yes(graph clustering) | No | No | No | Yes(visual attributes) | Yes(text processing) | No | No | No | No | Yes | No | No | No | No | No | No | Yes(User-study) | No | Yes(Singapore Foursquare) | None | No | No | No | No | No | No | None | No | No | Yes | No | No | Wikipedia, Tripadvisor | No | None | No | Foursquare: crawled from there mixing with other sources | ||||||||||||||||||||
115 | 2016 | Extension of RankGeoMF | Understanding the impact of weather for POI recommendations | Trattner, C., Oberegger, A., Eberhard, L., Parra, D., Marinho, L. | http://ceur-ws.org/Vol-1685/paper3.pdf | RecTour/RecSys | Conference | 1 | 8 | \cite{DBLP:conf/recsys/TrattnerOEPM16} | 8 | No | Yes | No | No | No | No | Stochastic Gradient Descent (SGD) | No | No | Yes | No | Yes(weather context is content) | No | No | Yes(Depending on RankGeo) | Yes | No | No | No | Yes(70% training, 10%validation, 20% test) | No | No | No | No | Yes(only US cities) | Yes. They show statistics of the cities (USA Cities Minneapolis, Boston, Miami, Honolulu) but reported the results of all together | Ranking | NDCG | No(RankGeoFm) | No | No | Yes(RankGeoFm) | Yes(10% validation) | Check-ins | No | No | Yes | No | No | No(They refer to Participatory cultural mapping based on collective behavior in location based social networks 2015) | Yes | Prev-No filtering | No | Foursquare: They refer to Participatory cultural mapping based on collective behavior in location based social networks 2015 | |||||||||||||||||||||
116 | 2016 | GeoSRS | GeoSRS: A hybrid social recommender system for geolocated data | Joan Capdevila, Marta Arias, Argimiro Arratia | Information Systems | Journal | 2 | 0 | \cite{DBLP:journals/is/CapdevilaAA16} | They say that they recommend locations but the training and test split are composed by tips | Yes(part of the hybrid approach) | Yes(LDA in part of the review content models) | Yes(LDA in part of the review content models) | No | No | Yes(CF + CB approach) | No | No | No | No | No | Yes(item features) | Yes(reviews) | No | No | Yes | No(as future work they want to do it) | No | Yes(70% of oldest tips to training) | No | No | No | No | No | No(but they clean the reviews analyzed) | Yes(Manhattan) | Ranking | AUC and Accuracy | No(TF-IDF, LDA) | No | No | No | No | POIs(Equivalent) | No | No | Yes | No | No | No | Not complete | None | No | Foursquare: no further information | ||||||||||||||||||||||
117 | 2016 | CPMFPPC | A hybrid method of recommending POIs based on context and personal preference confidence | Jian Li, Guanjun Liu, Changjun Jiang, ChunGang Yan | https://dl.acm.org/doi/abs/10.1145/3006299.3006330 | BDCAT | Conference | 3 | 0 | \cite{DBLP:conf/bdc/LiLJY16} | Yes(part of one of the probabilities) | Yes(part of the hybrid) | Yes(part of the hybrid approach) | No | No | Yes(two probabilities multiplied) | No | No | No | No | No | No | No | No | Yes | Yes | No | No | No | No | No | Yes(for each user 70% of the Check-ins to train rest to test) | No | No | Yes(removed first Check-in for every user, filter out by date and coordinates and removed users and items with less than 8 Check-ins) | Yes(Shanghai, filtering by coordinates range) | Ranking | Recall | No(LPR, PMF, CPMF, FMFMGM) | No | No | Yes(FMFMGM) | No | Check-ins | Yes. Figs 5-7 | No | No | No | SinaWeibo/DianPing | No | No | None | No | SinaWeibo: obtained from DianPing | ||||||||||||||||||||||
118 | 2016 | ---No-Acronym-- | Location, time, and preference aware restaurant recommendation method | Habib, Md.A., Rakib, Md.A., Hasan, M.A. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7860216 | ICCIT | Conference | ??? | 1 | 7 | \cite{Habib2016} | Only restaurant recommendation | No | No | No | No | No | Yes(Category, popularity, time awareness and distance) | No | No | No | Yes | No | Yes | No | No | Yes | Yes | No(although they claim to use an online recommender) | No | No information provided | No information provided | No information provided | No information provided | No information provided | No information provided | Yes(Only for Paris) | Yes(Paris) | None | None | Could no identify them | No | No | No | No | No information | No(although they discuss about it) | No | Yes | No | No | No(They refer to paper Participatory cultural mapping based on collective behavior in location based social networks 2015 and Nationtelescope: Monitoring and visualizing large-scale collective behavior in LBSNs 2015 | Yes | Post | No | Foursquare: Participatory cultural mapping based on collective behavior in location based social networks 2015 and Nationtelescope: Monitoring and visualizing large-scale collective behavior in LBSNs 2015 | ||||||||||||||||||||
119 | 2016 | TM-PFM | POI recommendation: A temporal matching between POI popularity and user regularity | Yao, Z., Fu, Y., Liu, B., Liu, Y., Xiong, H. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7837879 | ICDM | Conference | 1 | 39 | \cite{DBLP:conf/icdm/YaoFLLX16} | No | Yes (Probabilistic MF) | Yes (Probabilistic MF) | No | No | No | Maximum a Posteriori | No | No | No | No | Yes | No | No | Yes | Yes | No | No | No | No | Yes(80%-20% Check-in data) | No | No | No | Yes(users with less than 3 Check-ins removed) | Yes(New York) | Ranking | Precision, Recall, F1Measure | No(PMF, NMF,BPTF,LRT) | No | Yes(NMF, PMF) | No(They use temporal baselines but no geographical) | No | Check-ins | No | No | Yes | No | No | No(They refer to paper Modeling user activity preference by leveraging user spatial temporal characteristics in lbsns 2015) | Yes | Post | No | Foursquare: They refer to Modeling user activity preference by leveraging user spatial temporal characteristics in lbsns 2015 | ||||||||||||||||||||||
120 | 2016 | RCTF | Regularized content-Aware tensor factorization meets temporal-Aware location recommendation | Lian, D., Zhang, Z., Ge, Y., Zhang, F., Yuan, N.J., Xie, X. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7837944 | ICDM | Conference | 1 | 16 | \cite{DBLP:conf/icdm/LianZGZYX16} | No | Yes | Yes(They claim to follow a normal distribution... So according to our standards, yes) | No | No | No | Alternating Least Squares | No | No | Yes | No | No | No | No | Yes | Yes | No | No | No | Yes(for each user, 80% of the Check-ins for training, rest to test) | No | No | No | No | Yes(datasets prefiltered and locations visited by at least 5 users for Gowalla) | Yes(Jiepang, they say they craw from Beijing) | Ranking | Recall | No(UTE,WMF,CTMF-MTL,LRT,LTCR) | No | Yes(WMF) | Yes(LTCR) | No(held out validation) | Check-ins | No | Yes | No | No | Jiepang | No(They refer to Friendship and mobility: user move- ment in location-based social networks 2011) | Yes | Post | No | Gowalla and Jiepang: They refer to Friendship and mobility: user move- ment in location-based social networks 2011 for Gowalla | ||||||||||||||||||||||
121 | 2016 | IALBR | Interest aware location-based recommender system using geo-tagged social media | AlBanna, B., Sakr, M., Moussa, S., Moawad, I. | https://www.mdpi.com/2220-9964/5/12/245 | ISPRS | Journal | 1 | 7 | \cite{DBLP:journals/ijgi/AlbannaSMM16} | No | No | No | No | Yes | Yes(Interest and authority score) | No | No | No | Yes | No | Yes | No(They use tags as categories) | No | No | Yes(although it is strange) | No | No | No | No | No | No | No | No | Yes(Other mechanism. Queries for users within a radious 20km) | No | Yes(Only New York) | Ranking | Precision, Recall, F1 | Could no identify them | No | No | No | No | None | No | No | Yes | No | No | No(They refer to Participatory cultural mapping based on collective behavior data in location based social networks 2015 and Monitoring and visualizing large-scale collective behavior in lbsns 2011) | Yes | Prev-No filtering | No | Foursquare: They refer Participatory cultural mapping based on collective behavior data in location based social networks 2015 and Nationtelescope: Monitoring and visualizing large-scale collective behavior in lbsns 2015 | |||||||||||||||||||||
122 | 2016 | ---No-Acronym-- | Incorporating Spatial, Temporal, and Social Context in Recommendations for Location-Based Social Networks | Stepan, T., Morawski, J.M., Dick, S., Miller, J. | https://ieeexplore.ieee.org/document/7803653 | IEEE Transactions on Computational Social System | Journal | 1 | 28 | \cite{DBLP:journals/tcss/StepanMDM16} | Yes | No | No | No | No | Yes(Social, geographical, contextual) | No | No | No | Yes | Yes | No | No | No | Yes | Yes | No | No | Yes(80% of the first Check-ins for training, rest to test) | No | No | No | No | No | No | No | Ranking | Map, Precision, Recall, Top-1, Coverage | Could no identify them | No | No | No | No | Check-ins | No | Yes | Yes | No | Brightkite | No(They refer to Location-Based Social Network Data Repository 2014 and Friendship and mobility: User movement in location-based social networks 2012) | Yes | Prev-No filtering | No | Foursquare, Brightkite and Gowalla: http://www.public.asu.edu/~hgao16/ dataset.html for Foursquare and Friendship and mobility: User movement in location-based social networks 2012 for Gowalla and Brighkite | ||||||||||||||||||||||
123 | 2016 | ---No-Acronym-- | Time preference aware dynamic recommendation enhanced with location, social network and temporal information | Ozsoy, M.G., Polat, F., Alhajj, R. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7752347 | ASONAM | Conference | 1 | 4 | \cite{DBLP:conf/asunam/OzsoyPA16} | Yes | No | No | No | No | No | No | No | No | No(Hometown. Consider as geographical?) | Yes | No | No | No | Yes | Yes | No | No | Yes(January to training and February to test) | No | No | No | No | No | Yes(from the whole dataset they work only with 1 month for training and rest to test) | No | Ranking | Precision, NDCG, HitRate and coverage | No(CF, MO (with different families for each other)) | No | Yes(CF) | Yes(Hometown for CF and MO) | No | Check-ins | No | No | Yes | No | No | No(they refer to Multi-objective optimization based location and social network aware recommendation (2014)) | Yes | Post | No | Foursquare: They refer to Multi-objective optimization based location and social network aware recommendation 2014. Subset of Check-in2011 dataset | ||||||||||||||||||||||
124 | 2016 | GE | Learning graph-based poi embedding for location-based recommendation | Xie, M., Yin, H., Wang, H., Xu, F., Chen, W., Wang, S. | https://dl.acm.org/citation.cfm?id=2983711 | CIKM | Conference | 1 | 184 | \cite{DBLP:conf/cikm/XieYWXCW16} | Yes | No | Yes(Graph embedding) | Yes(Graph embedding) | No | No | Yes | Asynchronous Stochastic Gradient descent algorithm (ASGD) | No | Yes(graph embedding) | Yes | No | Yes | No(They say text, but are tags and categories) | Yes | Yes | Yes | No | No | No | Yes(80% of the Check-ins for training, 10% of train to validation and rest for test for every user) | No | No | No | No | No | Yes(USA for Foursquare) | Ranking | Accuracy(measured as Hit@k) | No(SVDFeature, JIM, PRME-G, Geo-Sage) | No | No | Yes(USG) | Yes | Check-ins | Yes(cold-start for the ones not having any check-in) | Yes | Yes | No | No | Yes(https://sites.google.com/site/dbhongzhi/) | Yes | Prev-No filtering | No(but in other paper, they refer to this URL) | Foursquare and Gowalla: Smae crawling strategy as Exploring millions of footprints in location sharing services 2011 | |||||||||||||||||||||
125 | 2016 | DeepReg | Regularising factorised models for venue recommendation using friends and their comments | Manotumruksa, J., MacDonald, C., Ounis, I. | https://dl.acm.org/citation.cfm?id=2983889 | CIKM | Conference | 1 | 16 | \cite{DBLP:conf/cikm/ManotumruksaMO16} | No(Although they use social regularization it is not collaborative) | Yes | No | No | No | No | Stochastic Gradient Descent | No | Yes(word embedding) | No | Yes(friends) | No | Yes(comments, word embedding) | No | No | Yes | No | No | No | No | No | No | No | Yes(5 fold cross-validation. 60% training, 20% validation and 20% test for every user) | No | No | Error | MAE, RMSE | No(MFN, MFP, MF, VMF, SoReg, BoWReg, SVD, TrustSVD) | No | Yes(MF) | No | Yes | POIs(Equivalent) | No | No | No | Yes | No | Yes(https://www.yelp.com/dataset/challenge) | Yes | Prev-No filtering | No | Yelp: https://www.yelp.com/dataset challenge | ||||||||||||||||||||||
126 | 2016 | ---No-Acronym--- | Context-aware location recommendations with tensor factorization | Zhu, X., Hao, R. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7636832 | CIC | Conference | 1 | 3 | \cite{DBLP:conf/iccchina/ZhuH16} | Time Aware Recommendation | No(They use a similarity matrix but not neighbours) | Yes | No | No | No | No | Stochastic Gradient Descent | No | No | No | No | Yes | No | No | Yes | Yes | No(although they claim to use an dynamic methods) | No | Yes(50% of the data as training) | No | No | No | No | No | Yes(users with mmore or equal 300 Check-ins and items with more or equals 100 visited) | Yes(New York) | Error | MAE, RMSE | No(Base, SVD++ time SVD++) | No | Yes(SVD++) | No | No | Check-ins | No | No | Yes | No | No | No(They refer to Modeling user activity preference by leveraging user spatial temporal characteristics in lbsns in 2015) | Not complete | None | No | Foursquare: They refer to Modeling user activity preference by leveraging user spatial temporal characteristics in lbsns 2015 | |||||||||||||||||||||
127 | 2016 | ---No-Acronym--- | Effective successive POI recommendation inferred with individual behavior and group preference | Chen, J., Li, X., Cheung, W.K., Li, K. | Neurocomputing | Journal | 1,2,3 | 21 | \cite{DBLP:journals/ijon/ChenLCL16} | It is group recommendation or at least they group users. Successive POI recommendation | No | Yes | Yes(Markov and BPR are used, I would say yes, although it is basic) | No | Yes(HITS) | No(If I understood correctly, they first perfporm a MF to predict the category and after that a graph-based method, but one after the other so it should not be considered as hybrid) | Group Bayesian Personalized Ranking BPR | Yes(grouping users) | No | Yes(Distances) | No | Yes(categories) | No | Yes(Markov Chains) | No | Yes | No | No | Yes(first 70% training and 30% test) | No | No | No | No | No | Yes(remove users with tips less than 50 and others where Check-in information cannot be located) | Yes(Los Angeles and New York) | Ranking | Precision | No(FPMC, FPMC-LR, MF) | No | Yes(MF) | No | No | POIs(Equivalent) | No | No | Yes | No | No | No(They refer to the paper: Location-based and preference-aware re- commendation using sparse geo-social networking data 2012) | Yes(both pre and post filtering) | Post | No | Foursquare: They refer to Location-based and preference-aware recommendation using sparse geo-social networking data 2012 | ||||||||||||||||||||||
128 | 2016 | STS_Grid, STS_DBSCAN | Location recommendation algorithm based on temporal and geographical similarity in location-based social networks | Yuan, Z., Li, H. | https://ieeexplore.ieee.org/document/7578804 | WCICA | Conference | ??? | 1 | 7 | \cite{Yuan2016} | Yes | No | No | No | No | No | No | Yes | No | Yes | No | No | No | No | Yes | Yes | No | No | Yes(each dataset is divided in the proportion of 8:2 in the time dimension) | No | No | No | No | No | No | Yes(Texas) | Ranking | Precision and RunTime | No(STS_DBSCAN) | No | No | No | No | Check-ins | No | Yes | No | No | No | No | Yes | Prev-No filtering | No | Gowalla: no further details | |||||||||||||||||||||
129 | 2016 | ---No-Acronym-- | Mining semantic location history for collaborative poi recommendation in online social networks | Pipanmekaporn, L., Kamolsantiroj, S. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7573686 | OBD | Conference | 1 | 4 | \cite{DBLP:conf/obd/PipanmaekapornK16} | Yes | No | No | No | No | No | No | No | No | Yes | No | Yes(categories) | No | No | No | Yes | No | No | No | No | No | No | No | No | Yes(They select 20 random users and hide some of the locations) | No | Yes(Tokyo and New York) | Ranking | Precision, Recall and F1 Measure | No(SVD-CF, LCF) | No | Yes(SVD) | No | No | POIs | No | No | Yes | No | No | No(They refer to Discovering and profiling overlapping communities in location-based social networks 2014) | Yes | Prev-No filtering | No | Foursquare: They refer to Discovering and profiling overlapping communities in location-based social networks 2014 | |||||||||||||||||||||
130 | 2016 | MAPS | MAPS: A multi aspect personalized POI recommender system | Baral, R., Li, T. | https://dl.acm.org/citation.cfm?id=2959187 | RecSys | Conference | 1,3 | 28 | \cite{DBLP:conf/recsys/BaralL16} | No | No | No | No | Yes(Topic Sensitive PageRank) | No | No | No | No | Yes(distance) | Yes(friends) | Yes(categories) | No | No | Yes | Yes | No | No | No | No | No | No | Yes(5-fold cross-validation CC) no constraints mentioned | No | No | No | Ranking | Precision, Recall, F1 | No(USG, LSBNRank and LFBCA) | No | No | Yes(USG) | No | Check-ins | No | Yes | No | No | Weeplaces | No(They refer to Personalized point-of-interest recommendation by mining users’ preference transition 2013) | Yes | Prev-No filtering | No | Weeplaces and Gowalla: Personalized point-of-interest recommendation by mining users’ preference transition 2013 from Gowalla | ||||||||||||||||||||||
131 | 2016 | BPRLR1, BPRLR2 | A unified point-of-interest recommendation framework in Location-based social networks | Cheng, C., Yang, H., King, I., Lyu, M.R. | https://dl.acm.org/citation.cfm?id=2901299 | TIST | Journal | 1,3 | 37 | \cite{DBLP:journals/tist/ChengYKL16} | No | Yes(MF) | Yes | No | No | Yes(MGM approach with BPR and Matrix factorization) | No | BPR | Yes(greedy clustering algorithm) | No | Yes | No | No | No | No | No | Yes | No | No | No | No | Yes(70% training and 30% of the observed data) | No | No | No | Yes(For Gowalla, users with less than 10 Check-ins removed and items with less than 20 visits. For Fousquare users with less than 10 Check-ins removed) | No | Ranking | Precision and Recall | No(MGM, PMF, PMFSR, PFM, FMFMGM, BPR, GeoMF, Rank-GeoFM) | No | Yes(BPR) | Yes(RankGeoFm, FMFMGM) | No | Check-ins (BUT in the training/test split they say that there is no overlap. However, no further info is provided) | Yes | Yes | Yes | No | No | No(For foursquare they refer to this paper: Exploring millions of footprints in location sharing services 2011) | Yes | Post | No | Foursquare and Gowalla: Exploring millions of footprints in location sharing services 2011 for Foursquare and no info for Gowalla | |||||||||||||||||||||
132 | 2016 | ---No-Acronym-- | Development of location-aware place recommendation system on Android smart phones | Jueajan, B., Naleg, K., Pipanmekaporn, L., Kamolsantiroj, S. | https://ieeexplore.ieee.org/document/7519252 | ICT-ISPC | Conference | ??? | 1 | 4 | \cite{Jueajan2016} | Yes | No | No | No | No | No | No | No | No | No | No | Yes | No | No | No | Yes | No | No | No | No | No | No | No | No | Yes(Select only 20 users and hide some of their visited locations) | No | No | Ranking | Precision and Recall | No | No | No | No | No | POIs | No | No | Yes | No | No | No | No | None | No | Foursquare: no more infor provided | ||||||||||||||||||||
133 | 2016 | PNS (first approach), CNF (second approach) | Location Recommendations for New Businesses Using Check-in Data | Eravci, B., Bulut, N., Etemoglu, C., Ferhatosmanoglu, H. | https://ieeexplore.ieee.org/document/7836791 | ICDM | Conference | 1 | 6 | \cite{DBLP:conf/icdm/EravciBEF16} | Only for business. Not generic POI. Propose 2 approaches | Yes(CNF) | No | Yes(PNS) | No | No | No | No | No | No | Yes(PNS, CNF) | No | Yes(PNS, CNF) | No | No | No | Yes | No | No | No | No | No | No | No | No | Yes(It is split by distance) | Yes(only POIs as business and removed POIs with less than 5) | Yes(New York) | Ranking | Accuracy | No(only the proposed method compared) | No | No | No | No | Check-ins | No | No | Yes | No | No | No | Not complete | None | No | Foursquare: I do not have the reference (the reported reference does not appear in the paper) | ||||||||||||||||||||
134 | 2016 | TICRec | TICRec: A probabilistic framework to utilize temporal influence correlations for time-aware location recommendations | Zhang, J.-D., Chow, C.-Y. | https://ieeexplore.ieee.org/document/7061519 | IEEE Transactions on Services Computing | Journal | 1 | 58 | \cite{DBLP:journals/tsc/ZhangC16} | Time Aware | Yes(they use the friends as collaborative) | No | Yes | No | No | No | No | No | No | Yes | Yes | No | No | No | Yes | Yes | No | No | Yes(half of the Check-ins with older timestamps to train rest to test) | No | No | No | No | No | No | No | Ranking | Precision and Recall | No(iGeoRec, LRT, UTE, GTAG) | No | No | Yes(GTAG, UTESE) | No | Check-ins | No | Yes | Yes | No | No | No(They refer to gSCorr: Modeling geo-social correla- tions for new check-ins on location-based social networks 2011 and Friendship and mobility: User movement in location-based social networks 2011) | Yes | Prev-No filtering | No | Foursquare and Gowalla: They refer to gSCorr: Modeling geo-social correla- tions for new check-ins on location-based social networks 2012 and Friendship and mobility: User movement in location-based social networks 2011 for Gowalla | |||||||||||||||||||||
135 | 2016 | USGT | Point-of-interest recommendation using temporal orientations of users and locations | Hosseini, S., Li, L.T. | https://link.springer.com/chapter/10.1007/978-3-319-32025-0_21 | DASFAA | Conference | 1 | 14 | \cite{DBLP:conf/dasfaa/HosseiniL16} | It seems that it is USG with a temporal modification | Yes | No | Yes(although they are simple, the authors state it is probabilistic) | No | No | Yes(KNN, SocialKNN and Geographical and Temporal) | Normal Equation (NE) | No | No | Yes | Yes | No | No | No | Yes | Yes | No | No | No | No | No | Yes | No | No | No | No | Ranking | Precision and Recall | No(UBCF, UBCFT, USGT) | No | Yes(UBCF) | Yes(USG) | No | POIs | No | No | Yes | No | Brightkite | Yes(http://www.public.asu.edu/∼hgao16/. but it does not work and https://snap.stanford.edu/data/loc-brightkite.html.) | Yes | Prev-No filtering | No | Foursquare and Brightkite: http://www.public.asu.edu/∼hgao16/ for Foursquare and https://snap.stanford.edu/data/loc-brightkite.html for Brighkite | |||||||||||||||||||||
136 | 2016 | CTS | CTS: Combine temporal influence and spatial influence for time-aware POI recommendation | Zhang, H., Yang, Y., Zhang, Z. | https://link.springer.com/chapter/10.1007/978-981-10-2053-7_25 | ICYCSEE | Conference | 1 | 3 | \cite{DBLP:conf/icycsee/ZhangYZ16a} | Time Aware | Yes | No | No | No | No | Yes(user collaborative filtering with geographical and temporal influences) | No | No | No | Yes | No | No | No | No | Yes | Yes | No | No | No | Yes(75% for training rest for test for every user of the most recent Check-ins) | No | No | No | No | Yes(only user with more than 10 Check-ins) | Yes(More less. Singapore Foursquare and California and Nevada for Gowalla) | Ranking | Precision and Recall | No(UCF, UCF+G, UTF, LRT, UTE+SE, BPP) | No | Yes(UCF) | Yes(UTESE) | No | Check-ins | No | Yes | Yes | No | No | No(They refer to papers Friendship and mobility: user movement in location- based social networks 2011 and GeoSoCa: exploiting geographical, social and categorical correlations for point-of-interest recommendations 2015 | Not complete | None | No | Foursquare and Gowalla: for Foursquare they refer to GeoSoCa: exploiting geographical, social and categorical correlations for point-of-interest recommendations 2015 and Friendship and mobility: user movement in location- based social networks 2011 for Gowalla | |||||||||||||||||||||
137 | 2016 | CoSoLoRec | CoSoLoRec: Joint factor model with content, social, location for heterogeneous point-of-interest recommendation | Guo, H., Li, X., He, M., Zhao, X., Liu, G., Xu, G. | https://link.springer.com/chapter/10.1007/978-3-319-47650-6_48 | KSEM | Conference | 1 | 5 | \cite{DBLP:conf/ksem/GuoLHZLX16} | Yes | Yes(Topic distribution + probabilistic latent factor) | Yes(Topic distribution + probabilistic latent factor) | No | No | Yes(geographical, textual, probabilistic, friend-based) | Maximum likelihood Estimation and Stochastic Gradient descent | No | No | Yes(distance) | Yes(friends) | No | Yes(topics and texts, topics extracted from text) | No | No | Yes | No | No | Yes(70%training according to review date) | No | No | No | No | No | Yes(users with more than 300 friends and less than 20 and for foursquare filter out user with more than 500 friends and less than 18) | Yes(More less Phoenix and Las Vegas and United States) | Error/Ranking | rPrecision and RMSE | No(PMF, NMF, BNMF, GT-BNMF, Geo-PFM,C-PMF, C-NMF, C-BNMF) | No | Yes(NMF) | Yes(Geo-PMF) | No | Check-ins | No | No | Yes | Yes | No | No | Yes | Post | No | Yelp and Foursquare: They refer to Learning from the crowd: regression discontinuity estimates of the effects of an online review database 2012 for Yelp | ||||||||||||||||||||||
138 | 2016 | PLTSRS | Preference-aware successive POI recommendation with spatial and temporal influence | Debnath, M., Tripathi, P.K., Elmasri, R. | https://link.springer.com/chapter/10.1007/978-3-319-47880-7_21 | SocInfo | Conference | 1 | 6 | \cite{DBLP:conf/socinfo/DebnathTE16} | Successive POI recommendation | Yes(part of the hybrid) | No | Yes(Markov) | No | No | Yes(Collaborative filtering, categories, temporal popularity) | No | No | No | No(It is only used as candidate) | No | Yes | No | Yes(Markov, successive) | Yes | Yes | No(although they claim to use an online recommender) | No | Yes(first 8 months training 1 month test) | No | No | No | No | No | Yes(removed POIs with less than 10 Check-ins) | Yes(Only New York) | Ranking | Precision and Recall | Yes(Pop, UCF, PCF, NMF) | Yes(Popularity) | Yes(UCF) | No | No | Check-ins | No | No | Yes | No | No | No | Not complete(they filter out POIs but I do not know if the users are the same) | Post | No | Foursquare: not more info | |||||||||||||||||||||
139 | 2016 | STPMF | A spatial-temporal probabilistic matrix factorization model for point-of-interest recommendation | Li, H., Hong, R., Wu, Z., Ge, Y. | https://epubs.siam.org/doi/10.1137/1.9781611974348.14 | SIAM | Conference | 1 | 12 | \cite{DBLP:conf/sdm/LiHWG16} | No | Yes(probabilistic matrix factorization) | Yes(probabilistic matrix factorization) | No | No | No | Gradient descent | No | No | Yes | No | Yes | No | No | Yes | Yes | No | No | No | Yes(80% earlier Check-ins to train, rest to test for every user) | No | No | No | No | Yes(removed users that have visited less than 5 lcoations and locations visited by less than 2 users) | No | Error/Ranking | Precision, Recall, NDCG and RMSE | No(NMF, PMF, UC, UG, LRT) | No | Yes(UCF, NMF) | Yes(UG) | No | POIs | No | No | Yes | No | No | No(they refer to Location-based social network data repository (2014) and Exploring millions of footprints in location sharing services (2011) | Yes | Possible post | No | Foursquare: They refer to Location-based social network data repository 2014 and Exploring millions of footprints in location sharing services 2011 | ||||||||||||||||||||||
140 | 2016 | Topical-GeoMF | A hybrid method for POI recommendation: Combining check-in count, geographical information and reviews | Xu, X., Zhao, P., Liu, G., Gu, C., Xu, J., Wu, J., Cui, Z. | https://link.springer.com/chapter/10.1007/978-3-319-45817-5_13 | APWeb | Conference | 1 | 0 | \cite{DBLP:conf/apweb/XuZLGXWC16} | No | Yes | No | No | No | No(They say that it is an hybrid approach but They are adding more things to the MF) | No | No | No | Yes | No | No | Yes(reviews) | No | No | Yes | No | No | No | No | No | Yes(randomly select x% of the visited locations of the user for training and 1-x to test) | No | No | Yes(removed users that have visited less than 10 locations) | No | Ranking | Precision and Recall | No(PMF, GeoMF) | No | No(PMF) | Yes(GeoMF) | No | POIs | No | No | Yes | No | No | No | Yes | Post | No | Foursquare: no further details | ||||||||||||||||||||||
141 | 2016 | CoTF | Context-aware point of interest recommendation using tensor factorization | Maroulis, S., Boutsis, I., Kalogeraki, V. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7840694 | BigData | Conference | 1 | 9 | \cite{DBLP:conf/bigdataconf/MaroulisBK16} | No | Yes | No | No | No | No | High Order Singular Value Decomposition (HOSVD), gradient descent | No | No | No | No | Yes | No | Yes(Transition, updated sequential pattern) | No | Yes | No | No | No | No | No | Yes(80% visiting locations to train 20% to test randomly FOR EVERY USER) | No | No | Yes(removed users that have visited less than 10 different locations and removed POIs visited by less than 2 users) | Yes(New York and Tokyo) | Ranking | Precision and Recall | No(WMF, GeoMF) | No | Yes(WMF) | Yes(GeoMF) | No | POIs | No | No | Yes | No | No | No(They refer to this paper: Modeling user activity preference by leveraging user spatial temporal characteristics in lbsns 2015) | Yes | Post | No | Foursquare: They refer to Modeling user activity pref- erence by leveraging user spatial temporal characteristics in lbsns 2015 | ||||||||||||||||||||||
142 | 2016 | MUG | Modeling user mobility via user psychological and geographical behaviors towards point of-interest recommendation | Chen, Y., Li, X., Li, L., Liu, G., Xu, G. | https://link.springer.com/chapter/10.1007/978-3-319-32025-0_23 | DASFAA | Conference | 1 | 1 | \cite{DBLP:conf/dasfaa/ChenLLLX16} | No | Yes(probabilistic MF) | Yes(power law, probabilistic MF) | No | No | Yes(Probabilistic and MF) | No | Yes | No | Yes | No | No | No | No | No | Yes | No | No | No | No | Yes(80%-20% partition datasets) | No | No | No | Yes(randomly selected a subset) | No | Ranking | Precision, Recall and MAP | No(MF,PMF,NMF,MGM) | No | Yes(MF) | Yes(MGM) | No | Check-ins | Yes(10% for training and rest to test) | Yes | No | No | Brightkite | Yes(http://snap.stanford.edu/data/loc-brightkite.html http://snap.stanford.edu/data/loc-gowalla.html) | Yes | Post | No | Gowalla and Brighkite: http://snap.stanford.edu/data/loc-gowalla.html for Gowalla and http://snap.stanford.edu/data/loc-brightkite.html for Brightkite | ||||||||||||||||||||||
143 | 2016 | ATTF | Aggregated Temporal Tensor Factorization Model for Point-of-interest Recommendation | Zhao, S., Lyu, M.R., King, I. | https://link.springer.com/chapter/10.1007/978-3-319-46675-0_49 | ICONIP | Conference | 1 | 7 | \cite{DBLP:conf/iconip/ZhaoLK16} | No | Yes | No(They use BPR for optimizing) | Yes | No | No | BPR | No | Yes(embedding neural network) | No | No | No | No | No | Yes | Yes | No | No | No | No | No | Yes(80%-20% of the Check-ins of each user) | No | No | Yes(removed POIs checked-in by less than 5 users and removed users with less than 10 Check-ins) | No | Ranking | Precision, Recall, F score | No(WRMF, BPR-MF, LRT, FPMC-LR) | No | Yes(BPRMF, WRMF) | No | No | Check-ins | No | Yes | Yes | No | No | No( they refer to paper: gscorr: modeling geo-social correlations for new check-ins on location-based social networks (2012) for foursquare and Capturing geographical influence in poi recommendations. In: International conference on neural information processing for Gowalla (2013)) | Yes | Post | No | Foursquare and Gowalla: They refer to gscorr: modeling geo-social correlations for new check-ins on location-based social networks 2012 for Foursquare and Capturing geographical influence in poi recommendations. In: Interna- tional conference on neural information processing 2013 for Gowalla | ||||||||||||||||||||||
144 | 2016 | TGTM-1,TGTM-2 | TGTM: Temporal-geographical topic model for point-of-interest recommendation | Zheng, C., Haihong, E., Song, M., Song, J. | https://link.springer.com/chapter/10.1007/978-3-319-32025-0_22 | DASFAA | Conference | 1 | 3 | \cite{DBLP:conf/dasfaa/ZhengESS16} | No | Yes(probabilistic matrix factorization) | Yes(probabilistic matrix factorization) | No | No | No | No | No | No | Yes | No | No | Yes | No | Yes | Yes | No | No | No | No | Yes(80%-20% of the dataset, although not specifically stated as random) | No | No | No | Yes(users with less than 10 Check-ins removed and each POI must b visited at least 10 times) | No | Error/Ranking | NMAE, NRMSE, Precision | No(BasicMF, GeoCF, UPT, TLA) | No | Yes(BasicMF) | Yes(GeoCF) | No | Check-ins | No | No | Yes | No | No | No(they refer to paper: Exploring millions of footprints in location sharing services (2011)) | Yes | Post | No | Foursquare and Gowalla: they refer to Exploring millions of footprints in location sharing services 2011 | ||||||||||||||||||||||
145 | 2016 | SSR | Point-of-Interest Recommendations via a Supervised Random Walk Algorithm | Xu, G., Fu, B., Gu, Y. | https://ieeexplore.ieee.org/abstract/document/7389906 | IEEE Intelligent Systems | Journal | 1 | 19 | \cite{DBLP:journals/expert/XuFG16} | No | No | No(They say something bout transition probabilities but they are not shown. I would vote for NO) | No | Yes | No | No | No | No | Yes | Yes | No | Yes | No | No | Yes | No | No(they claim to to a case study but I do not know how) | No | No | Yes(70% training and 30% test of the whole dataset) | No | No | No | Yes(filtered out data but no further info provided) | No | Ranking | Precision and HitRatio | No(CF, USG, RWR) | No | Yes(CF) | Yes(USG) | No | POIs(Equivalent) | No | No | Yes | No | No | No(They refer to this paper: Exploring Millions of Footprints in Location Sharing Services 2011 | Not complete (only Check-ins and users) | None | No | Foursquare: They refer to Exploring Millions of Footprints in Location Sharing Services 2011 | ||||||||||||||||||||||
146 | 2016 | MultiGran | Boosting point-of-interest recommendation with multigranular time representations | Rojas, G., Seco, D., Serrano, F. | http://www.jucs.org/jucs_22_8/boosting_point_of_interest | JUCS | Journal | 1 | 0 | \cite{DBLP:journals/jucs/RojasSS16} | Time Aware | Yes | No | No | No | No | No | No | No | No | No | No | No | No | No | Yes | Yes | No | No | Yes(along with random split) | No | Yes(along with CC temporal split) | No | No | No | Yes(removed users with less than 5 Check-ins and removed POIs with less than 5 Check-ins) | No | Ranking | Not any classic IR metric | No(They take the baseline of this paper: A Survey of Point-of-Interest Recommenda- tion in Location-Based Social Networks (2015)) | No | No | No | No | Check-ins | No | Yes | No | No | No | No(they refer to Friendship and mobility: user movement in location-based social networks 2011 | Not complete (Only stated the Check-ins) | None | No | Gowalla: They refer to Friendship and mobility: user movement in location-based social networks 2011 | |||||||||||||||||||||
147 | 2016 | SSR | Power of bosom friends, POI recommendation by learning preference of close friends and similar users | Fang, M.-Y., Dai, B.-R. | https://link.springer.com/chapter/10.1007/978-3-319-43946-4_12 | DaWaK | Conference | 1 | 4 | \cite{DBLP:conf/dawak/FangD16} | No | No | No | No | No | Yes(Different social models) | No | No | No | No | No | Yes | No | No | No | No | Yes | No | No | No information provided | No information provided | No information provided | No information provided | No information provided | No information provided | Yes(removed users that made less than 5 Check-ins and removed POIs which are visited by less than 5 users) | No | Ranking | Precision and Recall | No(USG) | No | No | Yes(USG) | No | No information | No | Yes | No | No | Brightkite | No(they refer to paper Friendship and mobility: user movement in location-based social networks 2011 | Not complete (Only Check-ins) | None | No | Gowalla and Brightkite: They refer to Friendship and mobility: user movement in location-based social networks 2011 | |||||||||||||||||||||
148 | 2016 | SGMF | Point of interest recommendation with social and geographical influence | Zhang, D.-C., Li, M., Wang, C.-D. | https://ieeexplore.ieee.org/document/7840709 | BigData | Conference | 1 | 12 | \cite{DBLP:conf/bigdataconf/ZhangLW16} | Yes | No | No(They say probability but i think it is too obvious) | No | No | Yes(Social, KNN and geographical component) | No | No | No | Yes | Yes | No | No | No | No | Yes | No | No | No | No | Yes(80% training - 20% test) | No | No | No | Yes(remove users with less thatn 10 Check-ins and POIs with less than 5 visits) | Yes(New York and Washington) | Ranking | Precision and Recall | No(UBCF, SBCF, GI,US-BCF, GM-FCF) | No | Yes(UBCF) | Yes(GM-FCF) | No | Check-ins | No | Yes | No | No | No | No | Yes | Post | No | Gowalla: They refer to Fused matrix factor- ization with geographical and social influence in location-based social networks 2012 | ||||||||||||||||||||||
149 | 2016 | GeoTeCS | GeoTeCS: Exploiting geographical, temporal, categorical and social aspects for personalized poi recommendation | Baral, R., Wang, D., Li, T., Chen, S.-C. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7785729 | IRI | Conference | 1,3 | 20 | \cite{DBLP:conf/iri/BaralWLC16} | Time Aware | No | Yes | Yes(They use estimation of probability) | No | No | No | Gradient descent | No | No | Yes(distance) | Yes(friends) | Yes | No | No | Yes | Yes | No | No | No | No | No | No | Yes | No | No | No | Ranking | Precision and Recall and F-score | No(GeoMFTD, FMFMGM, USG) | No | No | Yes(USG, FMFMGM) | No | No information | No | Yes | No | No | Weeplaces | No(they refer to other paper: Personalized point-of-interest recommendation by mining users’ preference transition 2013) | Yes | Prev-No filtering | No | Weeplaces and Gowalla: Personalized point-of-interest recommendation by mining users’ preference transition 2013 for Gowalla | |||||||||||||||||||||
150 | 2016 | RC, DCC | A recommender system research based on location-based social networks | Wang, J., Tan, R., Zhang, R.-P., You, F. | https://link.springer.com/chapter/10.1007/978-3-319-39910-2_8 | SCSM | Conference | 1 | 3 | \cite{DBLP:conf/hci/WangTZY16} | Yes | No | No | No | No | No | Yes. Maybe we should categorize it as other type of algorithm. Clustering for both | No | Yes | No | Yes | No | Yes | No | No | No | Yes | No | No | No | No | No | Yes(80% Check-ins training 20% test for every user) | No | No | No | No | Ranking | Precision and Recall | None | No | No | No | No | Check-ins | No | No | No | No | Sina microblog app | No | No | None | No | Sina microblog app: no further details | |||||||||||||||||||||
151 | 2016 | GME, GME-S | Graph-based metric embedding for next POI recommendation | Xie, M., Yin, H., Xu, F., Wang, H., Zhou, J. | https://link.springer.com/chapter/10.1007/978-3-319-48743-4_17 | WISE | Conference | 1 | 11 | \cite{DBLP:conf/wise/XieYXWZ16} | Next POI recommendation | No | Yes(Graph embedding) | Yes(Graph embedding) | No | No | No | asynchronous stochastic gradient algorithm (ASGD) | No | Yes(graph based embedding) | No | No | No | No | Yes(GME-S) | Yes(for GME-S I think yes as they exploit a threshold but for classic GME) | Yes | No | No | No | Yes(weird split but create sequences and 20% of the ratings to test) | No | No | No | No | No | No(although they say the users live in California, Check-ins are around the world) | Ranking | Hit@k | No(PRME, SPORE, BPR) | No | Yes(BPR) | No | No | Check-ins | No | No | Yes | No | Yes(https://sites.google.com/site/dbhongzhi/) | Yes | Prev-No filtering | No | Foursquare and Twitter: https://sites.google.com/site/dbhongzhi/. Their twitter dataset is a Foursquare dataset | ||||||||||||||||||||||
152 | 2016 | ---No-Acronym-- | Location based recommender system using enhanced random walk model | Katarya, R., Ranjan, M., Verma, O.P. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7913179 | PDGC | Conference | ??? | 1 | 1 | \cite{Katarya2016} | No | No | No | No | Yes | No | No | No | No | No | No | No | No | No | No | Yes | No | No | No | No | Yes(not correctly specified) | No | No | No | No | No | Ranking | Precision and Recall | Yes(Pop) | Yes(Pop) | No | No | No | POIs | No | Yes | Yes | No | Brightkite | Yes(https://snap.stanford.edu/data/loc-brightkite.html for Brightkite but the rest are not correctly processed | Not complete (only Check-ins) | None | No | Foursquare, Gowalla and Brightkite: they refer to the standford links (but are not correctly copied) and they also provide an url for Fourquare but it is not correctly copied | |||||||||||||||||||||
153 | 2016 | ---No-Acronym-- (They define as novel Scheme) | Location privacy protected recommendation system in mobile cloud | Guan, H., Qian, H., Zhao, Y. | https://link.springer.com/chapter/10.1007/978-3-319-48671-0_36 | ICCCS | Conference | 1 | 0 | \cite{DBLP:conf/icccsec/GuanQZ16} | Time Aware | Yes | No | Yes(I would say yes, Kernel density estimation) | No | No | Yes(Categories, geographical and friends) | No | No | No | Yes | Yes | Yes | No | No | Yes | Yes | No | No | No Information | No Information | No Information | No Information | No Information | No Information | No | No | Ranking | Precision and Recall | No(GeoSoca, Core, GeoMF, sPCLR) | No | No | Yes(Geosoca, GeoMF) | No | No information | No | Yes | No | No | No | No | No | None | No | Gowalla: not further details | |||||||||||||||||||||
154 | 2016 | ---No-Acronym-- | Personalized location recommendations with local feature awareness | Zhu, X., Hao, R., Chi, H., Du, X. | https://ieeexplore.ieee.org/document/7842140 | GLOBECOM | Conference | 1 | 4 | \cite{DBLP:conf/globecom/ZhuHCD16} | No | Yes(Labeled-LDA) | Yes(Labeled-LDA) | No | No | Yes(User preference minin and local feature interference) | No | No | No | No | No | Yes | No | No | No | Yes | No | No | No | No | No | No | No | No | Yes(A city selected as test set, rest as training) | No | No | Ranking | Precision and Recall | No(LPA, UPAR, LFAR) | No | No | No | No | No information | No | No | Yes | No | No | No(They refer to this paper: Lcars: a location-content- aware recommender system 2013 | Not complete (only number of users and records) | None | No | Foursquare: they refer to Lcars: a location-content- aware recommender system 2013 | |||||||||||||||||||||
155 | 2016 | ASMF | Point-of-interest recommendations: Learning potential check-ins from friends | Li, H., Ge, Y., Hong, R., Zhu, H. | https://dl.acm.org/doi/10.1145/2939672.2939767 | SIGKDD | Conference | 1,3 | 172 | \cite{DBLP:conf/kdd/LiGHZ16} | Yes | No | Yes | No | No | Yes(They talk about Random-Walk but not clear how they use it) | No | Alternate Least Squares | No | No | Yes | Yes | Yes | No | No | No | Yes | No | No | No | Yes(80% more ancient for each user for train rest 20% to test) | No | No | No | No | Yes(for no cold-start evaluation, removed users and items with less than 10 locations visites/users that have visited) | Yes(California for Foursquare) | Ranking | Precision, Recall and MAP | No(WMF, USG, IRENMF, BPR, PMF, RegPMF, other cominations of ASMF) | No | Yes(BPR, MF) | Yes(USG, IRENMF) | No | POIs, Specifically stated that they aggregate the interactions | Yes(address new item and new user recommendations) | Yes | Yes | No | No | No | Yes | Post | No | Gowalla and Foursquare: no further details given | |||||||||||||||||||||
156 | 2016 | STELLAR | STELLAR: Spatial-temporal latent ranking for successive point-of-interest recommendation | Zhao, S., Zhao, T., Yang, H., Lyu, M.R., King, I. | http://www.aaai.org/ocs/index.php/AAAI/AAAI16/paper/view/12249 | AAAI | Conference | 1 | 127 | \cite{DBLP:conf/aaai/ZhaoZYLK16} | Successive POI recommendation | Yes | No | Yes | No | No | No | No | Pairwise Ranking model | No | No | No(in the other survey they claim to use it, but i do not know how) | No | No | No | No | Yes | Yes | No | No | Yes | No | No | No | No | No | Yes(Remove POIs with less than 5 different users checked in and users who checked-in more than 10 times) | No | Ranking | Precision and Recall | No(BPRMF, WRMF, LRT, FPMC, TLAR, SLAR) | No | Yes(MF, BPR) | No(Depends in TLSAR AND SLAR use it, but it seems no) | No | Check-ins | No | Yes | Yes | No | No | No(They refer to other papers: gSCorr: Modeling geo-social correlations for new check-ins on location-based social networks (2012), Capturing geo- graphical influence in POI recommendations (2013)) | Yes | Post | No | Foursquare and Gowalla: they refer to gSCorr: Modeling geo-social correlations for new check-ins on location-based social networks (2012) for Foursquare and Capturing geographical influence in POI recommendations (2013) for Gowalla | ||||||||||||||||||||
157 | 2016 | ---No-Acronym-- | Inferring a personalized next point-of-interest recommendation model with latent behavior patterns | He, J., Li, X., Liao, L., Song, D., Cheung, W.K. | http://www.aaai.org/ocs/index.php/AAAI/AAAI16/paper/view/12361 | AAAI | Conference | 1 | 85 | \cite{DBLP:conf/aaai/HeLLSC16} | Next POI recommendation | Yes | No | Yes(probabilistic MF and FPMC) | Yes(probabilistic MF and FPMC) | No | No | No | BPR-Expectation Maximization | No | No | Yes(geographical in the MF transition) | No | No | No | Yes | No | Yes | No | No | No Information | No Information | No Information | No Information | No Information | No Information | Yes(Remove POIs with less than 5 different users checked in and users who checked-in more than 10 times) | Yes(Foursquare is Los ANgeles. Gowalla is full) | Ranking | Precision | No(MF, PMF, FPMC-LR) | No | Yes(MF) | No | No | No Information | No | Yes | Yes | No | No | No(They refer to: Location- based and preference-aware recommendation using sparse geo-social networking data 2012 and Fused ma- trix factorization with geographical and social influence in location-based social networks. 2012 | Yes | Post | No | Foursquare and Gowalla: they refer to Location- based and preference-aware recommendation using sparse geo-social networking data 2012 for Foursquare and Fused ma- trix factorization with geographical and social influence in location-based social networks. 2012 for Gowalla | ||||||||||||||||||||
158 | 2016 | WWO | Unified point-of-interest recommendation with temporal interval assessment | Liu, Y., Liu, C., Liu, B., Qu, M., Xiong, H. | https://dl.acm.org/doi/10.1145/2939672.2939773 | SIGKDD | Conference | 1 | 73 | \cite{DBLP:conf/kdd/LiuLLQX16} | Time Aware | Yes | No | No | Yes | No | No | No | No | Yes | No | No | No | No | No | Yes | Yes | Yes | No | No | Yes(80% more recent to train 20% newest test) | No | No | No | No | No | Yes(remove users and items with less that 10 ratings, for gowalla) | No | Ranking | Precision, Recall, F0.5, NDCG | No(FPMC, PMF, FPMC-LR, pimf) | No | No(PMF) | No | No | Check-ins | No | Yes | Yes | No | No | Yes(https://snap.stanford.edu/data/loc-gowalla.html Only Gowalla) | Yes | Post | No | Foursquare and Gowalla: they refer to https://snap.stanford.edu/data/loc-gowalla.html for Gowalla. No info for Foursquare | ||||||||||||||||||||
159 | 2016 | ELR-DC | ELR-DC: An Efficient Recommendation Scheme for Location Based Social Networks | Lv, R., Wang, Y., Jin, Q., Ma, J. | https://ieeexplore.ieee.org/document/7917155 | iThings, GreenCom, CPSCom, SmartData | Conference | 1 | 1 | \cite{DBLP:conf/ithings/LvWJM16} | Yes | No | No | No | No | No | No | No | No | No | Yes | No | No | No | No | Yes | No | No | No | No | Yes(80% training 20% test) | No | No | No | Yes(Users with more than 30 friends and 20 Check-ins. Each location must be checked more than 10 times) | No | Ranking | Precision, Recall | No(FCF, CCF, CFCF) | No | Yes(CCF) | No(not sure, i think not) | No | No information (I think it is by Check-ins) but Not sure (they use the number of time the user have checked in a POI) | No | Yes | No | No | No | No | Not complete (only Check-ins, nodes and edges) | None | No | Gowalla: no further details | ||||||||||||||||||||||
160 | 2016 | GeoBPR | Joint geo-spatial preference and pairwise ranking for point-of-interest recommendation | Yuan, F., Jose, J.M., Guo, G., Chen, L., Yu, H., Alkhawaldeh, R.S. | https://ieeexplore.ieee.org/document/7814578 | ICTAI | Conference | 1 | 27 | \cite{DBLP:conf/ictai/YuanJGCYA16} | No | Yes | Yes(It uses BPR) | No | No | No | BPR, Stochastic Gradient Descent | No | No | Yes | No | No | No | No | No | Yes | No | No | No | No | No | No | Yes (5 cross-validation at dataset level) | No | Yes(users with less than 20 ratings and items with less than 5) | Yes (Phoenix and Las Vegas) | Ranking | Precision and Recall | Yes (Most popular, UB, BPR) | Yes(Pop) | Yes(UCF) | Yes(FMF) | Use cross valiadation | POIs(Equivalent) | No | No | No | Yes | No | Yes(https://www.yelp.co.uk/dataset/challenge) | Yes | Post | No | Yelp: www.yelp.co.uk/dataset challenge | ||||||||||||||||||||||
161 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
162 | 2017 | ??? | Rational interest spots recommendation dependent on location-based user self expression features | Li, P. | ??? | ??? | ??? | ??? | 1 | 0 | IGNORE | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
163 | 2017 | ??? | Social media mining and visualization for point-of-interest recommendation | Xingyi, R., Meina, S., Haihong, E., Junde, S. | ??? | ??? | ??? | ??? | 1,2 | 2 | IGNORE | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
164 | 2017 | ??? | Chapter 5 - Smart cities, urban sensing, and big data: mining geo-location in social networks | D.Sacco, G.Motta, L.-l.You, N.Bertolazzo, F.Carini, T.-y.Ma | https://www.sciencedirect.com/science/article/pii/B9780128120132000058 | Big Data and Smart Service Systems | Journal | ??? | 2 | 0 | IGNORE | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
165 | 2017 | ??? | Exploiting temporal influence for point-of-interest recommendation | Oppokhonov, S., Park, S. | ??? | ??? | ??? | ??? | 1 | 0 | IGNORE | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
166 | 2017 | ---No-Acronym-- | Privacy preserving location recommendations | Badsha, S., Yi, X., Khalil, I., Liu, D., Nepal, S., Bertino, E. | WISE | Conference | 1 | 8 | Not on the scope. No POI recommendation approach but criptography one. IGNORE | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
167 | 2017 | ---No-Acronym-- | Point-Of-Interest Recommender System for Social Groups | Gottapu, R.D., Sriram Monangi, L.V. | https://www.sciencedirect.com/science/article/pii/S1877050917318148 | Procedia Computer Science | Conference | ??? | 1,2,3 | 5 | I think it is not on the scope. It is recommendation for groups. IGNORE | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
168 | 2017 | ---No-Acronym-- | A study on recommendation systems in location based social networking | Kullappa, L.S., Kumar, R.A., Kullappa, R. | https://jios.foi.hr/index.php/jios/article/view/1071 | Journal of Information and Organizational Science | Journal | ??? | 1 | 0 | It is a survey, not proposing any model. IGNORE | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
169 | 2017 | FRPC-A | Friend recommendation considering preference coverage in location-based social networks | Yu, F., Che, N., Li, Z., Li, K., Jiang, S. | https://link.springer.com/chapter/10.1007/978-3-319-57529-2_8 | PAKDD | Conference | 1 | 23 | Friend recommendation. IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
170 | 2017 | GSD-PPG | Personalized POI groups recommendation in location-based social networks | Yu, F., Li, Z., Jiang, S., Yang, X. | https://link.springer.com/chapter/10.1007/978-3-319-63564-4_9 | APWeb-WAIM | Conference | 1 | 2 | Maybe is not on the scope. POI group recommendation. IGNORE --> although it recommends POI it is focused on groups of POIs | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
171 | 2017 | ---No-Acronym-- | Community detection and location recommendation based on LBSN | Su, C., Jia, X.-T., Xie, X.-Z., Li, N. | https://ieeexplore.ieee.org/document/8842736 | ICNISC | Conference | ??? | 1 | 1 | I think it is not on the scope. Community detection, not POI recommendation. IGNORE | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
172 | 2017 | ---No-Acronym-- | Research on group POIs recommendation fusion of users' gregariousness and activity in LBSN | Yuan, Z., Chen, C. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7951929 | ICCCBDA | Conference | ??? | 1 | 4 | I think it is not on the scope. Group recommendation. IGNORE | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
173 | 2017 | GGR | Where could we go? Recommendations for groups in location-based social networks | Ayala-Gomez, F., Daróczy, B., Mathioudakis, M., Benczúr, A., Gionis, A. | https://dl.acm.org/citation.cfm?doid=3091478.3091485 | WebSci | Conference | 1,3 | 11 | I think it is not on the scope. Group recommendation. IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
174 | 2017 | ---No-Acronym-- | Deep collaborative filtering approaches for context-Aware Venue Recommendation | Manotumruksa, J. | https://dl.acm.org/citation.cfm?id=3084159 | SIGIR | Conference | 1 | 3 | No approach provided. Only 1 page. Doctoral consortium. IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
175 | 2017 | ??? | A fuzzy collaboration system for ubiquitous loading/unloading space recommendation in the logistics industry | Toly Chen, Chi-Wei Lin | https://www.sciencedirect.com/science/article/pii/S0736584516300825 | Robotics and Computer-Integrated Manufacturing | Journal | ??? | 2 | 0 | Logistics, not POI recommendation. IGNORE | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
176 | 2017 | ??? | Indigenization of urban mobility | Zimo Yang, Defu Lian, Nicholas Jing Yuan, Xing Xie, Yong Rui, Tao Zhou, | https://www.sciencedirect.com/science/article/pii/S0378437116309062 | Physica A: Statistical Mechanics and its Applications | Journal | ???- There is a URL of this paper but one of archiv of 2014 https://dblp.uni-trier.de/rec/bibtex/journals/corr/YangLYXRZ14 | 2 | 0 | It does not propose a specific POI recommendation approach but a method to select algorithms if the user are locals or tourists. IGNORE | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
177 | 2017 | POILP | Point-of-interest recommendation for location promotion in location-based social networks | Yu, F., Li, Z., Jiang, S., Lin, S. | https://ieeexplore.ieee.org/abstract/document/7962475 | MDM | Conference | 1 | 5 | \cite{DBLP:conf/mdm/YuLJL17} | POI Recommendation with location promotion. I think they evaluate only for categories, so should we remove it? Finally IGNORE | No | No | Yes | No | No | No | No | No | No | Yes | Yes(friends of the user) | Yes | No | No | No | Yes | No | No | No Information | No Information | No Information | No Information | No Information | No Information | No | No | Ranking | Precision, Recall | No(CKNN, USG, UP-Based) | No | No | Yes(USG) | No | No information | No | Yes | Yes | No | No | No | Not complete | None | No | Foursquare and Gowalla: No further information | |||||||||||||||||||||
178 | 2017 | TBLR | Location recommendation algorithm for online social networks based on location trust | Lei, B., Zhanquan, W., Sun, H., Huang, S. | https://ieeexplore.ieee.org/document/8298763 | EIIS | Conference | ??? | 1 | 0 | ??? | I will say that this papers should be removed from the survey. The paper is bad writter and the format is awful. IGNORE | Yes | No | No | No | No | Yes(Probabilistic and using similarity between users) | No | Yes | No | No | No | No | No | No | No | Yes | No | No | No | No | No | No | No | No | Yes(Weird evaluation. It seems they only select one user as target user) | No | No | Ranking | Precision and Recall | No(only variations of the model) | No | No | No | No | Check-ins | No | No | No | No | No | No | No | None | No | ??? | |||||||||||||||||||
179 | 2017 | LURWA | Personalized location recommendation by aggregating multiple recommenders in diversity | Lu, Z., Wang, H., Mamoulis, N., Tu, W., Cheung, D.W. | https://link.springer.com/article/10.1007/s10707-017-0298-x | GeoInformatica | Journal | 1 | 16 | \cite{DBLP:journals/geoinformatica/LuWMTC17} | Repeated in 2015. Maybe it is not in the scope. IGNORE. The same as \cite{DBLP:conf/recsys/LuWMTC15} | No | No | No | No | No | Yes(combines different location recommenders) | No | No(used in evaluation, not in the method) | No | No | No | No | No | No | Yes | Yes | No | No | Yes([1, t -deltat], for training, (t-deltat, t] for validation and (t, t+ deltat) for test | No | No | No | No | No | No(although they define the active users) | No | Ranking | Precision, Recall and Utility | No(USG, iGSLR, RankBoost, BPRMF, LRT, SBPT, GeoMF) | No | Yes(CF) | Yes(GCF) | Yes | Check-ins | No | Yes | Yes | No | No | Yes | No | Post | No | Foursquare and Gowalla: not further information | |||||||||||||||||||||
180 | 2017 | ---No-Acronym-- | Selecting and weighting users in collaborative filtering-based POI recommendation | Ríos, C., Schiaffino, S., Godoy, D. | https://ri.conicet.gov.ar/handle/11336/59727 | Acta Polytechnica Hungarica | Journal | ??? | 1 | 1 | \cite{Rios2017} | Maybe we should remove it. They do not propose a specific algorithm and It is not published anywhere--> they compare different ways to select user neighbors for POI recommendation | Yes | No | No | No | No | No | No | No | No | Yes | Yes | Yes | Yes(process tips) | No | No | Yes | No | No | No | No | Yes(70% training, 30% test) | No | No | No | Yes(Only New York) | Yes(New York) | Error | MAE | No(Classsic UB) | No | Yes(UB) | No | No | POIs | No | No | Yes | No | No | No(They refer to Location-based and preference-aware recommendation using sparse geo-social networking data 2012) | Not complete | None | No | Foursquare: Location-based and preference-aware rec- ommendation using sparse geo-social networking data 2012 | ||||||||||||||||||||
181 | 2017 | USTTc (for POI recommendation) | A time-aware spatio-textual recommender system | Pavlos Kefalas, Yannis Manolopoulos | https://www.sciencedirect.com/science/article/abs/pii/S095741741730009X | Expert Systems with Applications | Journal | 2 | 0 | \cite{DBLP:journals/eswa/KefalasM17} | They propose 2 algorithms. One for review recommendation and one for POI recommendation (the one being analyzed) | Yes(part of the hybrid approach) | No | No | No | No | Yes(spatial and social) | No | No | No | Yes | No | No | Yes(textual) | No | No(only in the splitting method) | Yes | No | No | No | Yes(For each user, two sets, the training and the test one by Check-ins. Experiment repeated 30 times but also temporal analysis) | No | Yes(For each user, two sets, the training and the test one by Check-ins. Experiment repeated 30 times but also temporal analysis) | No | No | No | Yes(They indicate that the dataset contain information about these cities: Charlotte, Edinburgh, Karlsruhe, Las Vegas, Madison, Montreal, Phoenix, Pitts- burgh, Urbana-Champaign and Waterloo, but they do not split the dataset in the cities) | Ranking | Pecision and Recall | No(U, US, UT, UST, UTF, UTE+SE) | No | Yes(U) | Yes(UTESE) | No | Check-ins | No | No | No | Yes | No | Yes(https://www.yelp.com/dataset/challenge) | Yes | Prev-No filtering | No | Yelp: https://www.yelp.com/dataset _ challenge and they also refer to other papers | |||||||||||||||||||||
182 | 2017 | Geo-Teaser | Geo-teaser: Geo-temporal sequential embedding rank for point-of-interest recommendation | Zhao, S., Zhao, T., King, I., Lyu, M.R. | https://dl.acm.org/citation.cfm?doid=3041021.3054138 | WWW | Conference | 1 | 96 | \cite{DBLP:conf/www/ZhaoZKL17} | Yes | No | Yes(seems PMF) | Yes(seems PMF, althout it is very simple) | Yes(skipgram) | No | No(They say it is a combination of 2 models, but it seems that the the score formula uses the second model) | Stochastic Gradient Descent | No | Yes(word2vec, skip-gram) | Yes | No | No | No | Yes | Yes | Yes | No | No | No | Yes(for each user first 80% as training, 20% as test) | No | No | No | No | Yes(Pois checked by less than 5 users and users whith less than 10 Check-ins removed) | No | Ranking | Pecision and Recall | No(BPRMF, WRMF, LRT, LORE, Rank-GeoFM, SG-CWARP) | No | Yes(BPRMF, MF) | Yes(RankGeoFm) | No | Check-ins | No | Yes | Yes | No | No | No(They refer to gSCorr: Modeling geo-social correlations for new check-ins on location-based social networks 2012 for Foursquare and Capturing geographical influence in POI recommendations 2013 for Gowalla) | Yes | Post | Yes(https://github.com/henryslzhao/geo_teaser) | Yes | No | Yes(for their approach) | Foursquare and Gowalla: they refer to gSCorr: Modeling geo-social correlations for new check-ins on location-based social networks 2012 for Foursquare and Capturing geographical influence in POI recommendations 2013 for Gowalla | ||||||||||||||||||
183 | 2017 | HGMF | Exploiting hierarchical structures for POI recommendation | Zhao, P., Xu, X., Liu, Y., Zhou, Z., Zheng, K., Sheng, V.S., Xiong, H. | https://ieeexplore.ieee.org/document/8215538 | ICDM | Conference | 1 | 17 | \cite{DBLP:conf/icdm/ZhaoXLZ0SX17} | No | Yes | No | No | No | No | Projected gradient descent | No | No | Yes | No | No | No | No | No | Yes | No | No | No | No | No | Yes(for each user, 30% of the user Check-ins to test) | No | No | Yes(filtered pois with less than two visitors) | Yes(Nevada, California and Singapore) | Ranking | Precision and Recall | No(UCF, UCF+G, GeoMF, HSR) | No | Yes(UCF) | Yes(GeoMF) | No(although they select some parameters by validation) | Check-ins | No | Yes | Yes | No | No | No(They refer to Time-aware point-of-interest recommendation 2013) | Yes | Post | No | Foursquare and Gowalla: They refer to Time-aware point-of-interest recommendation 2013 | ||||||||||||||||||||||
184 | 2017 | PRFMC | A personalised ranking framework with multiple sampling criteria for venue recommendation | Manotumruksa, J., Macdonald, C., Ounis, I. | https://dl.acm.org/doi/10.1145/3132847.3132985 | CIKM | Conference | 1 | 19 | \cite{DBLP:conf/cikm/ManotumruksaMO17a} | No | Yes | Yes | No | No | Yes(it combines several actions. See Equation 5 of the paper) | BPR, Gradient descent | Yes | No | Yes | Yes | No | No | No | No | Yes | No | No | No | No | No | No | Yes(5 fold-cross validation at dataset level) | No | Yes(remove users and venues with less than 10 interactions) | No | Ranking | MAP, NDCG, MRR | No(MGM, GBPR, PRFMC, SPLD, SBPR.SWBPR, PRFMC, GeoSo, GSBPR, BPR, BPRMC, PRMFC) | No | Yes(BPR) | Yes(GeoBPR) | Yes(5 fold validation with specifically stated validation percentage) | Check-ins (But they aggregate the number of times the user visited a venue) | No | Yes | No | Yes | Brightkite | Yes(https://snap.stanford.edu/data/ for Gowalla and Birghtkite and https://www.yelp.com/dataset challenge for Yelp) | Yes | Post | No | Gowalla, Brightkite and Yelp: https://snap.stanford.edu/data/ for Gowalla and Brighkite and https://www.yelp.com/dataset challenge for Yelp | ||||||||||||||||||||||
185 | 2017 | GRMF and MLRP | A deep recurrent collaborative filtering framework for venue recommendation | Manotumruksa, J., Macdonald, C., Ounis, I. | https://dl.acm.org/doi/10.1145/3132847.3133036 | CIKM | Conference | 1 | 31 | \cite{DBLP:conf/cikm/ManotumruksaMO17} | This is a special case, because the both approaches are concatented, they are not independent | No | Yes | No | Yes | No | Yes(it users the factors as input of the neural network) | No | No | No | Yes | No | No | No | Yes | No(I think not, only sequential) | Yes | No | No | No | Yes(Most recent rating as test, then select other 100 venues that has not visited before and perform the evaluation, leave one out) | No | No | No | No | Yes(remove venues with less than 10 Check-ins) | No | Ranking | HR and NDCG | No(MF, BPR, GeoBPR, RNN, DREAM, NeoMF. GMF, MLP) | No | Yes(BPR) | Yes(GeoBPR) | No | Check-ins | Yes(cold start users = users with less than 10 Check-ins) | No | Yes | Yes | Brightkite | Yes(https://archive.org/details/201309_foursquare_dataset_umn for foursquare, https://snap.stanford.edu/data/ for Brightkite and https://www.yelp.com/dataset/challenge for Yelp) | Yes | Post | No(Although he gives the code for NeuroMF, a baseline https://github.com/hexiangnan/neural_collaborative_filtering) | Foursquare, Brightkite and Yelp: https://snap.stanford.edu/data for Brightkite, https://archive.org/details/201309 foursquare dataset umn for Foursquare and https://www.yelp.com/dataset challenge for Yelp | |||||||||||||||||||||
186 | 2017 | SPTW | A Reliable Point of Interest Recommendation based on Trust Relevancy between Users | Logesh, R., Subramaniyaswamy, V. | https://link.springer.com/article/10.1007/s11277-017-4633-1 | Wireless Personal Communications | Journal | 1,3 | 30 | \cite{DBLP:journals/wpc/RaviS17} | No | Yes | No | No | Yes(Random Walk) | No | No | No | No | No | No | Yes | No | No | No | Yes | No | No | No Information | No Information | No Information | No Information | No Information | No Information | No | No | Error/Ranking | RMSE, Coverage, Precision, FMeasure | No(TidalTrust, MoleTrust, TrustWalker, RelevantTrustWalker, SocialPertinentTrustWalker) | No | No | No | No | No information | No | Yes | Yes | No | Brightkite and Jiepang | No | Yes | Prev-No filtering | No | Foursquare, Gowalla, Jiepang, Brightkite but not further information provided. More users than check-ins. Weird | ||||||||||||||||||||||
187 | 2017 | TPR-UM | Exploiting User Mobility for Time-aware POI Recommendation in Social Networks | Zheng, H., Zhou, Y., Liang, N., Xiao, X., Sangaiah, A.K., Zhao., C. | https://ieeexplore.ieee.org/document/8082789 | IEEE Access | Journal | ??? | 1 | 6 | \cite{Zheng2017} | Time Aware | No | Yes(They claim probabilistic and it is clear, but should also be considered as factorization. I would vote for yes but it is 90% probabilistic) | Yes | No | No | No | No | Yes | No | Yes | No | No | No | No | Yes | Yes | No | No | No | No | Yes(70% of the dataset for training, 10% for validation and 20% to test) | No | No | No | Yes(remove users whose Check-ins is less than 10 and the POIs visited less than 5 times) | Yes(Foursquare uses Singapore) | Ranking | Precision and Recall, F1 measure | No(UTE, Geo-PFM, GAGT, TICRec) | No | No | Yes(GeoPFM) | Yes(10% of the data) | Check-ins | No | Yes | Yes | No | No | Yes(https://www.ntu.edu.sg/home/gaocong/datacode.htm and https://snap.stanford.edu/data/) | Yes | Post | No | Foursquare and Gowalla: https://snap.stanford.edu/data/ for Gowalla and http://www.ntu.edu.sg/home/gaocong/datacode.htm for Foursquare | ||||||||||||||||||||
188 | 2017 | UGSE-LR | A Grid-Based successive point-of-interest recommendation method | Gau, H.-Y., Lu, Y.-S., Huang, J.-L. | https://ieeexplore.ieee.org/document/8074153 | Ubi-Media | Conference | ??? | 1 | 3 | \cite{Gau2017} | Successive POI recommendation | Yes | No | No | No | Yes(Page Rank) | Yes(Similar to USG. Combines Region, CF and transition) | No | No | No | Yes | No | No | No | Yes | No | Yes | No | No | No | No | Yes(70% of the dataset for training, 10% for validation and 20% to test) However, not specifically stated that it was random | No | No | No | Yes(remove POIs checked by less than 80 users and remove users with less than 5 Check-ins) | No | Ranking | Precision and Recall | No(U, U-LR, UPW-LR, UG-LR, FPMC, FPMC-LR) | No | Yes(U) | Yes(UG-LR) | Yes(10% of the data) | Check-ins | No | Yes | No | No | Brightkite | Yes(http://snap.stanford.edu/data) | Yes | Post | No | Brightkite and Gowalla: http://snap.stanford.edu/data | ||||||||||||||||||||
189 | 2017 | VRer(ESSVM-UCP) | VRer: Context-Based Venue Recommendation using embedded space ranking SVM in location-based social network | Xia, B., Ni, Z., Li, T., Li, Q., Zhou, Q. | https://www.sciencedirect.com/science/article/pii/S0957417417302634 | Expert Systems with Applications | Journal | 1,2 | 15 | \cite{DBLP:journals/eswa/XiaNLLZ17} | Time Aware | No | No(They use SVMs, other) | No | No | No | No | Yes(SVMs I Would say yes) | No | No | Yes(It is not wor2vec but it is a kind of embedding for the SVM) | No | No | Yes | No | No | Yes(I think yes, but I'm not sure how do they use it) | Yes | No | No | No Information | No Information | No Information | No Information | No Information | No Information | No | Yes(Manhattan) | Ranking | Coverage, Precision, Recall, Popularity | Yes(UserCF, VenueCF, PoV, NNR) | Yes | Yes(NNR) | No | No | No information | No(They discuss about it) | No | Yes | No | No | No | Not complete (only users and tweets) | None | No | Foursquare: crawled using Twitter | ||||||||||||||||||||
190 | 2017 | TDLDA, TATD (both using the UPOST scheme) | Modeling User Preferences on Spatiotemporal Topics for Point-of-Interest Recommendation | Yang, S., Huang, G., Xiang, Y., Zhou, X., Chi, C.-H. | https://ieeexplore.ieee.org/document/8034986 | SCC | Conference | 1 | 3 | \cite{DBLP:conf/IEEEscc/YangHXZC17} | No | Yes(LDA) | Yes(LDA) | No | No | No | No | No | No | No(only for evaluation, they use the distance) | No | No | Yes | No | Yes | Yes | No | No | No | No | Yes(80% if the dataset for training, rest for test) | No | No | No | No | Yes(New York) | Ranking | Accuracy | No(USTTM) | No | No | Yes(USTTM is geographical) | No | Check-ins | No | No | Yes | No | No | No(They refer to Modeling user activity preference by leveraging user spatial temporal characteristics in lbsns 2015) | Yes | Prev-No filtering | No | Foursquare: they refer to Modeling user activity preference by leveraging user spatial temporal characteristics in lbsns 2015 | ||||||||||||||||||||||
191 | 2017 | CLB | Current location-based next POI recommendation | Oppokhonov, S., Park, S., Ampomah, I.K.E. | https://dl.acm.org/citation.cfm?id=3106528 | WI | Conference | 1 | 9 | \cite{DBLP:conf/webi/OppokhonovPA17} | Next POI recommendation | Yes | No | No | No | No | No, although they take into account sequential information apart from CF | No | No | No | No(They claim to use it but I do not see where) | No | No | No | Yes | Yes | Yes | No | No | No | No | No | Yes(for every user, 70% of his Check-ins for training and 30% for test) | No | No | Yes(remove POIs that was checked in by less than 10 users and remove users with less than 10 Check-ins) | Yes(New York and Tokyo) | Ranking | Precision and Recall | No(U, UTS) | No | Yes(UB) | Yes(UB + Geo) | No | Check-ins | No(stated as something to analyze for future work) | Yes | Yes | No | No | No | Yes | Post | No | Gowalla and Foursquare: no further details | |||||||||||||||||||||
192 | 2017 | ---No-Acronym-- | Using function approximation for personalized point-of-interest recommendation | Chen, B., Yu, S., Tang, J., He, M., Zeng, Y. | https://www.sciencedirect.com/science/article/pii/S0957417417300544 | Expert Systems with Applications | Journal | 1,2 | 7 | \cite{DBLP:journals/eswa/ChenYTHZ17} | Yes | No | No | No | No | Yes | Yes(approximation method. Chebyshev) | No | Yes | No | No | No | Yes | No | No | Yes | Yes | No | No | No | No | No | Yes(80% training, 20% test for each user) | No | No | Yes | Yes(Austin, Singapore, Stockholm, San Francisco, Dallas) | Error/Ranking | Precision, Recall, Doversity, MAE | No | No | No | No | No | POIs | No | Yes | Yes | No | No | No | Yes | Post | No | Foursquare and Gowalla: Foursquare they refer to Time-aware point-of-interest recommendation 2013 and Gowalla they refer to Friendship and mobility: User move- ment in location-based social networks 2011 | |||||||||||||||||||||
193 | 2017 | PACE | Bridging collaborative filtering and semi-supervised learning: A neural approach for POI recommendation | Yang, C., Bai, L., Zhang, C., Yuan, Q., Han, J. | https://dl.acm.org/doi/10.1145/3097983.3098094 | SIGKDD | Conference | 1,3 | 153 | \cite{DBLP:conf/kdd/YangBZY017} | Yes | No | No | No | Yes | No | No | Stochastic Gradient Descent | No | No(use embeddings but in DNN) | Yes | Yes | No | No | No | No | Yes | No | No | No | Yes(For each user 80% Check-ins for training rest to test) | No | No | No | No | Yes(For gowalla, filtered out users with less than 15 Check-ins and pois with less than 10 visitors. For Yelp, users with less than 10 POIs and POIs with less than 10 visitors, removed) | No | Ranking | Precision, Recall, NDCG and MAP | No(IrenMF, LOCABAL, USG, iGSLR, LORE, ASMF, ARMF) | No | No | Yes(IRENMF, LORE) | No | Check-ins | No | Yes | No | Yes | No | Yes for Yelp (https://www.yelp.com/dataset challenge) for Gowalla they refer to Friendship and mobility: user movement in location-based social networks 2011 | Yes | Prev | Yes(https://github.com/yangji9181/PACE2017) | Yes | No | No | Gowalla and Yelp: Friendship and mobility: user movement in location-based social networks 2011 for Gowalla and https://www.yelp.com/dataset challenge for Yelp | ||||||||||||||||||
194 | 2017 | LSARS | A location-sentiment-aware recommender system for both home-town and out-of-town users | Wang, H., Fu, Y., Wang, Q., Yin, H., Du, C., Xiong, H. | https://dl.acm.org/citation.cfm?id=3098122 | SIGKDD | Conference | 1 | 35 | \cite{DBLP:conf/kdd/WangFWYDX17} | No | Yes(They claim probabilistic and it is clear, but should also be considered as factorization as use latent variables. I would vote for yes but it is 90% probabilistic) | Yes | No | No | No | No | Yes | No | Yes | No | Yes | Yes | No | No | Yes | No | No | No | No | Yes | No | No | No | No | Yes(Foursquare, New York and Los Angeles and Yelp. 4 cities although no further info provided) | Ranking | Accuracy (defined a hit@k), and Precision | No(JIM, CAPRF, LCA-LDA, CKNN, UPS-CF) | No | No | Yes(JIM, UPS-CF) | No | POIs | No | No | Yes | Yes | No | Yes(https : //www.yelp.com/dataset challenge for yelp and http : //www.public.asu.edu/hgao16/dataset .html for foursquare. THe second one does not work) | Yes | Prev-No filtering | No | Foursquare and Yelp: | http : //www.public.asu.edu/hдao16/dataset.html for Foursquare and https://www.yelp.com/dataset challenge for Yelp | |||||||||||||||||||||
195 | 2017 | CTF-ARA | CTF-ARA: An adaptive method for POI recommendation based on check-in and temporal features | Si, Y., Zhang, F., Liu, W. | https://www.sciencedirect.com/science/article/pii/S0950705117301879 | Knowledge Based Systems | Journal | 1,2 | 28 | \cite{DBLP:journals/kbs/SiZL17} | Yes | No | No | No | No | No | No | Yes | No | No | No | No | No | Yes(consecutive. It is not specifically transition and so on, but i would vote for yes) | Yes | Yes | No | No | No | No | No | Yes(for every user, 16% of visited POIs to test, rest to train) | No | No | No | No. Although they refer to UTESE and I know they filter by city there (Singapore for Foursquare) | Ranking | Precision, Recall | No(U, LRT, UTE, UTE+FSUA, GTAG-BPP) | No | Yes(U) | Yes(GTAG) | No | POIs | No(They discuss a little about it) | Yes | Yes | No | No | No(They refer to Time-aware point-of-inter- est recommendation 2016) | Yes | Prev-No filtering | No | Foursquare and Gowalla: They refer to Time-aware point-of-inter- est recommendation 2013 | ||||||||||||||||||||||
196 | 2017 | AGSG | Aspect-aware point-of-interest recommendation with geo-social influence | Guo, Q., Sun, Z., Zhang, J., Chen, Q., Theng, Y.-L. | https://dl.acm.org/citation.cfm?id=3099066 | UMAP | Conference | 1 | 11 | \cite{DBLP:conf/um/GuoSZCT17} | No | No | No | No | Yes(PageRank) | No | No | No | No | Yes | Yes | Yes | Yes | No | No | Yes | No | No | No | No | No | No | Yes(5 fold cross validation at dataset level) | No | Yes(Users and items with less than 10 reviews, removed) | Yes(Phoenix, Las Vegas, Charlotte) | Ranking | Precision, Recall | No(Random, UCF, ICF, PMF, ItemRank, PPR, LFBCA, TriRank) | Yes(random) | Yes(UCF, ICF) | Yes(LFBCA) | No(cross-validation) | POIs(Equivalent) | No | No | No | Yes | No | Yes(https://www.yelp.com/dataset_challenge) | Yes | Prev-No filtering | No | Yelp: https://www.yelp.com/dataset_challenge | ||||||||||||||||||||||
197 | 2017 | ---No-Acronym-- | Location Recommendation Based on Social Trust | Wagih, H.M., Mokhtar, H.M.O., Ghoniemy, S.S. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8265107 | SKG | Conference | 1 | 3 | \cite{DBLP:conf/skg/WagihMG17} | No | No | No | No | Yes | No | No | No | No | No | Yes | No | No | No | No | Yes | No | No | No Information | No Information | No Information | No Information | No Information | No Information | Yes(Removed friends with no further connection) | No | Ranking | Accuracy (no further details provided) | No(CLoRW and the one from All friends are not equal: Using weights in social graphs to improve search (2010)) | No | No | No | No | POIs, Check-ins | No | Yes | No | No | Brightkite | No | Yes | Both | No | Gowalla and Brightkite: https://snap.stanford.edu/data/ | ||||||||||||||||||||||
198 | 2017 | PCTF | Partition-based collaborative tensor factorization for POI recommendation | Luan, W., Liu, G., Jiang, C., Qi, L. | https://ieeexplore.ieee.org/document/7974891 | Journal of Automatica Sinica | Journal | 1 | 33 | \cite{DBLP:journals/ieeejas/LuanLJQ17} | The pdf takes too long to load. Time-Aware | No | Yes | No | No | No | No | Element-wise gradient descent | Yes | No | No | No | Yes | No | No | Yes | Yes | No | No | No | No | Yes(10 groups of experiments and each group contains 10 sub-experiments, randomly choosing 70% of data for training and 30% for "validation") | No | No | No | Yes(users with more than 500 Check-ins and POIs with more than 50 Check-ins) | Yes(Shanghai) | Error | MAE, RMSE | No(TMF, CTF) | No | No | No | No | Check-ins | No | No | No | No | No | Dianping and Weibo | Yes | Post | No | Weibo and Dianping: they use Weibo and Dianping for expliting the characteristics | |||||||||||||||||||||
199 | 2017 | TAP | A temporal-aware POI recommendation system using context-aware tensor decomposition and weighted HITS | Ying, Y., Chen, L., Chen, G. | https://www.sciencedirect.com/science/article/pii/S0925231217303910 | Neurocomputing | Journal | 1,2,3 | 21 | \cite{DBLP:journals/ijon/YingCC17} | Time Aware | No | Yes(first part) | No | No | Yes(HITS-Based) | No | No | No | No | No | Yes(TAP-F) | Yes | No(I will say no, altought they claim to exploit tips) | No | Yes | Yes | No | No | No | No | No | Yes(70% location records as training 10% validation and 20% test) | No | No | Yes(User with more than 24 tips and 8 tips in quering city. Users from NYC) | Yes(NYC) | Error/Ranking | Precision, Recall, MAE, RMSE | No(LP-CF, TD-CF, HITS) | No | No | No | Yes | POIs | No | No | Yes | No | No | No | Yes | Prev | No | Foursquare: no further information | |||||||||||||||||||||
200 | 2017 | TGSC-PMF | Context-aware probabilistic matrix factorization modeling for point-of-interest recommendation | Ren, X., Song, M., Haihong, E., Song, J. | https://www.sciencedirect.com/science/article/pii/S0925231217302758 | Neurocomputing | Journal | 1,2,3 | 62 | \cite{DBLP:journals/ijon/RenSES17} | Yes | No | Yes(probabilistic MF) | Yes(probabilistic MF) | No | No | No(Probabilistic MF with social, categorical and geographical, I would vote for NO, as the hybrid part is integrated in a PMF so the final model is not hybrid. No as it is an aggregated model) | No | No | No | Yes | Yes | Yes | Yes | No | No | Yes | No | No | No | No | Yes(80% for training and 20% for test) | No | No | No | Yes(users need to have made more than 10 Check-ins and each POI should be visited 10 times at least. Users should have visited 5 different POIs) | No | Error/Ranking | Precision, Recall, MAE, RMSE | No(LCARS, TL-PMF, ACTF, USG, ASMF, NCPD) | No | No | Yes(USG, ASMF) | No | Check-ins | No | No | Yes | No | No(They refer to paper Exploring millions of footprints in location sharing services 2011) | Yes | Prev | No | Foursquare: They refer to Exploring millions of footprints in location sharing services 2011 for FOursquare and Twitter. THeir twitter dataset is also from Foursquare | ||||||||||||||||||||||
201 | 2017 | SEM-PPA | SEM-PPA: A semantical pattern and preference-aware service mining method for personalized point of interest recommendation | Zhu, L., Xu, C., Guan, J., Zhang, H. | https://www.sciencedirect.com/science/article/pii/S1084804516303502 | Journal of Network and Computer Applications | Journal | 1,2 | 15 | \cite{DBLP:journals/jnca/ZhuXGZ17} | No | No | No | No | Yes(They say they use HITS method) | No | No | Yes | No | Yes | Yes | Yes | No(semantic, but they are categories) | Yes | No | Yes | No | No | No | Yes(First half of historical trajetories. ) | No | No | No | No | Yes(Filtering ut latitude and longitude) | Yes(Beijing and some Europe and USA) | Ranking | Precision, Recall, F1 | No(PMF, PMFSR, LPCF, SEM-PPA) | No | No(PMF) | No | No | Check-ins | No | No | No | No | Geolife | No | No | None | No | Geolife: no further information | ||||||||||||||||||||||
202 | 2017 | ---No-Acronym-- | Exploiting location significance and user authority for point-of-interest recommendation | Yu, Y., Wang, H., Sun, S., Gao, Y. | https://link.springer.com/chapter/10.1007/978-3-319-57529-2_10 | PAKDD | Conference | 1 | 2 | \cite{DBLP:conf/pakdd/YuWSG17} | No | No | Yes | No | Yes(They discuss and say their approach is also based on PageRank and HITS) | No | Stochastic Gradient Descent | No | No | No | No | No | No | No(Markov Chain but no in temporal manner) | No | Yes | No | No | No | No | No | No | No | Yes(5 fold validation for each user) | No | No. Although they refer to UTESE and I know they filter by city there (Singapore for Foursquare) | Ranking | Precision and Recall | No(UserKnn, ItemKnn, PMF, WRMF,BPR-MF,GeoMF,PMF) | No | Yes(UB, IB) | Yes(GeoMF) | No(cross-validation) | POIs | No | Yes | Yes | No | No | Yes(https://www.ntu.edu.sg/home/gaocong/datacode.htm) | Yes | Prev-No filtering | No | Foursquare and Gowalla: http://www.ntu.edu.sg/home/gaocong/datacode.htm for both | ||||||||||||||||||||||
203 | 2017 | EIUCF, EIICF | Learning recency and inferring associations in location based social network for emotion induced point-of-interest recommendation | Logesh, R., Subramaniyaswamy, V. | Journal of Information Science and EngineerinG | Journal | 1 | 22 | \cite{DBLP:journals/jise/RaviS17} | This is a special case, because the both approaches are concatented, they are not independent models | Yes | No | No | No | No | Yes(They propose two CF approaches and then an hybrid one combining both of them) | No | No | No | No | No | Yes | No | No | No | Yes | No | No | No | No | Yes(from 1% to 60% Check-ins to train) | No | No | No | Yes(No information provided) | No | Ranking | HitRate, Recall, Precision, F1 | No(Random, BPR, TBCF, UCF, ICF) | Yes(Random) | Yes(BPR, UB, IB) | No | No | POIs(Equivalent) | No | No | No | Yes | TripAdvisor | No | Yes | Post | No | Yelp and Tripadvisor: no further information | ||||||||||||||||||||||
204 | 2017 | ---No-Acronym-- | Familiarity-aware POI recommendation in urban neighborhoods | Han, J., Yamana, H. | https://www.jstage.jst.go.jp/article/ipsjjip/25/0/25_386/_article | Journal of Information Processing | Journal | 1 | 1 | \cite{DBLP:journals/jip/HanY17} | No | Yes(They say they use LDA... For me, it would be 90% probabilistic and 10% factorization, but just because they use topic modeling) | Yes | No | No | No | No | Yes | No | Yes(They discuss about familiarty areas) | No | Yes | No | No | No | Yes | No | No | No | No | No | No | Yes(5 x 2 cross validation at dataset level) | No | Yes(Only users that have at least one activity area in the evaluation city) | Yes(New York, Los Angeles, Tokyo) | Ranking | Recall | No(LCARS, FamLCARS, Prop_NoFam, PropPrefFam) | No | No | Yes(GeoSage) | No(cross-validation) | Check-ins | No(They discuss a little about it ) | No | Yes | No | No | No | Yes | Post | No | Foursquare: they crawled the data using Twitter | ||||||||||||||||||||||
205 | 2017 | LBPR | Category-aware next point-of-interest recommendation via listwise Bayesian personalized ranking | He, J., Li, X., Liao, L. | https://www.ijcai.org/Proceedings/2017/255 | IJCAI | Conference | 1 | 45 | \cite{DBLP:conf/ijcai/HeLL17} | Next POI recommendation | Yes | No | Yes(They discuss about a descomposition) | Yes(BPR and other optimization, power law for geographical) | No | No | Yes, (geographical and categorical influence) | Similar to BPR | No | No | Yes | No | Yes | No | Yes | No | Yes | No | No | No | Yes(80% of the Check-ins for training rest to test) | No | No | No | No | No | Yes(New York and Los Angeles) | Ranking | Precision (by categories) and Precision for POI | No(MF, PMF, FPMC, PRME) | No | Yes(MF) | Yes(PRMGE) | No | Check-ins (But they say non-overlapping) | No | No | Yes | No | No | No | Yes | Prev-No filtering | No | Foursquare: no further information | ||||||||||||||||||||
206 | 2017 | IEMF | Learning user's intrinsic and extrinsic interests for point-of-interest recommendation: A unified approach | Li, H., Ge, Y., Lian, D., Liu, H. | https://www.ijcai.org/Proceedings/2017/294 | IJCAI | Conference | 1 | 18 | \cite{DBLP:conf/ijcai/LiGLL17} | No | Yes | No | No | No | No | Stochastic Gradient Descent (SGD) using | Yes | No | Yes | No | No | No | No | No | Yes | No | No | No | Yes(80% train 20% test for every user) | No | No | No | No | No | No | Ranking | Precision, Recall and MAP | No(ARMF, IrenMF, USG, BPR, WRMF) | No | No | Yes(IRENMF) | No | Check-ins (but specifically stated that they aggregate the Check-ins) | No | Yes | Yes | No | No | No | Yes | Prev-No filtering | No | Gowalla and Foursquare: no further information | ||||||||||||||||||||||
207 | 2017 | Geo-PRMF | Geo-pairwise ranking matrix factorization model for point-of-interest recommendation | Zhao, S., King, I., Lyu, M.R. | https://link.springer.com/chapter/10.1007/978-3-319-70139-4_37 | ICONIP | Conference | 1 | 8 | \cite{DBLP:conf/iconip/ZhaoKL17} | No | Yes | No | No | No | No | BPR | No | No | Yes | No | No | No | No | No | Yes | No | No | No | No | No | Yes(80% of each users Check-ins as training data rest to test) | No | No | Yes(POIs with more o equal 5 Check-ins and users that have checked at least 20 times) | No | Ranking | Precision and Recall | No(BiasedMF, BPR-MF, MGMMF,GeoMF) | No | Yes(BPRMF) | Yes(GeoMF) | No | Check-ins | No | Yes | Yes | No | No | No(They refer to Exploring social-historical ties on location-based social networks 2012 and Friendship and mobility: user movement in location-based social networks 2011) | Yes | Post | No | Foursquare and Gowalla: they refer Exploring social-historical ties on location-based social networks 2012 for Foursquare and Friendship and mobility: user movement in location-based social networks 2011 for Gowalla | ||||||||||||||||||||||
208 | 2017 | VPOI | What your images reveal: Exploiting visual contents for point-of-interest recommendation | Wang, S., Wang, Y., Tang, J., Shu, K., Ranganath, S., Liu, H. | https://dl.acm.org/doi/10.1145/3038912.3052638 | WWW | Conference | 1 | 118 | \cite{DBLP:conf/www/WangWTSRL17} | Yes | No | Yes(probabilistic MF) | Yes(probabilistic MF) | Yes | No | No | Gradient descent | No | No | No | No | Yes(images of POIs) | No | No | No | Yes | No | No | No | No | No | Yes(20-40% of the POIs for training For each user, rest to test) | No | No | Yes(filtering images tagged as selfies and locations that hae been visited by just 1 user and users with at least 8 visited locations) | Yes(New York and Chicago) | Ranking | Precision and Recall | No(UCF, VUCF, NMF, PMF, VBPR) | No | Yes(UCF) | No | No(cross-validation) | POIs | Yes(for them, a cold-start user is ne user with no Check-ins in the training set) | No | No | No | No | Yes | Post | No | Instagram: crawled from Instagram API | ||||||||||||||||||||||
209 | 2017 | LBA and BF | Behavior-based location recommendation on location-based social networks | Rahimi, S.M., Wang, X., Far, B. | https://link.springer.com/chapter/10.1007/978-3-319-57529-2_22 | PAKDD | Conference | 1 | 5 | \cite{DBLP:conf/pakdd/RahimiWF17} | No | Yes(BF, LBA) | Yes(LBA, BF) | No | No | Yes(for the BF, takes into account the probability of the user in location with respect his behavior, but no for LBDA) | Gradient descent for BF, Expectation maximization LBA | No | No | Yes | No | Yes | No | No | Yes(modelized as behavior) | Yes | No | No | No | No | No | Yes(1 random Check-in for each user. Leave one out) | No | No | No | No | Ranking | Precision and Recall | No(USG, PMM) | No | No | Yes(USG) | No | Check-ins | Yes(cold start are users with less than five Check-ins in the training dataset) | Yes | Yes | No | Brightkite | No(They refer to A study of recommending locations on location-based social network by collaborative filtering (2012)) | Yes | Prev-No filtering | No | Gowalla: They refer to A study of recommending locations on location-based social network by collaborative filtering 2012 | ||||||||||||||||||||||
210 | 2017 | CBGeoMFC | Geographical and overlapping community modeling based on business circles for POI recommendation | Li, M.-R., Huang, L., Wang, C.-D. | https://link.springer.com/chapter/10.1007/978-3-319-67777-4_60 | IScIDE | Conference | 1 | 3 | \cite{DBLP:conf/iscide/LiHW17} | Yes(Pearson correlation for similariy matrix but no neighbourhood) | Yes | Yes(They say the use normal distribution for geographical information, KDE) | No | No | No | Stochastic Gradient Descent (SGD) using | Yes | No | Yes | No | Yes | No | No | No | Yes | No | No | No | No | Yes(80% for training and 20% for test after agreggating) | No | No | No | No | Yes(New York for Foursquare and Hong Kong from Jiepang) | Ranking | Precision, Recall and F1 | No(UCF, PMF, MFC) | No | Yes(UB) | No | No | Check-ins (But specifically states that they aggregate the Check-ins) | No | No | Yes | No | Jiepang | No(They refer to : GeoMF: joint geographical modeling and matrix factorization for point-of-interest recommendation 2014 and Modeling user activity preference by leveraging user spatial temporal characteristics in LBSNs 2015) | Yes | Prev-No filtering | No | Foursquare and Jiepang: GeoMF: joint geographical modeling and matrix factorization for point-of-interest recommendation 2014 and Modeling user activity preference by leveraging user spatial temporal characteristics in LBSNs 2015 but not sure which paper is used in each dataset | ||||||||||||||||||||||
211 | 2017 | NH-JTI | Jointly modeling heterogeneous temporal properties in location recommendation | Hosseini, S., Yin, H., Zhang, M., Zhou, X., Sadiq, S. | https://link.springer.com/chapter/10.1007/978-3-319-55753-3_31 | DASFAA | Conference | 1 | 6 | \cite{DBLP:conf/dasfaa/HosseiniYZZS17} | Yes(their model is based on CF ) | No | Yes | No | No | No | Expectation maximization, Normal Equation | Yes | No | No | No | No | No | No | Yes(temporal decay) | Yes | No | No | No | No | No | Yes(30% of visited POIs for the user) | No | No | No | No | Ranking | Precision, Recall and F1 | No(CF,CFT,USG,USGT) | No | Yes(CF) | Yes(USG) | No | POIs | No(they discuss about it but no specific experiments are performed) | No | Yes | No | Brightkite | No(they refer to Point-of-interest recommendation using temporal orientations of users and locations 2016) | Yes | Prev-No filtering | No | Foursquare and Brighkite: They refer to Point-of-interest recommendation using temporal orienta- tions of users and locations 2016 for both | ||||||||||||||||||||||
212 | 2017 | SCCF | Social and Content Based Collaborative Filtering for Point-of-Interest Recommendations | Xu, Y.-N., Xu, L., Huang, L., Wang, C.-D. | https://link.springer.com/chapter/10.1007/978-3-319-70139-4_5#Abs1 | ICONIP | Conference | 1 | 2 | \cite{DBLP:conf/iconip/XuXHW17} | No | Yes | No | No | No | Yes(Social and MF using Categorical information) | Alternate Least Squares | No | No | No | Yes | Yes | No(They claim to use reviews but they do not process texts) | No | No | Yes | No | No | No Information | No Information | No Information | No Information | No Information | No Information | Yes(social constraints and at least 15 check-ins records) | Yes (different geographical locations and states) | Ranking | Precision, Recall, F1 | No | No | No | No | No | No Information | Yes(address new item and new user recommendations) | No | No | Yes | No | Yes (for yelp it seems we need to put our name and email) | Not complete (number of Check-ins not stated) | Post | No | Yelp: no further details | ||||||||||||||||||||||
213 | 2017 | ARNN | Attention-based recurrent neural network for location recommendation | Xia, B., Li, Y., Li, Q., Li, T. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8258747 | ISKE | Conference | 1 | 15 | \cite{DBLP:conf/iske/XiaLLL17} | No | No | No | Yes | No | No | No | No | No | Yes | No | Yes | No | Yes | No | Yes | No | No | Yes(split the data in different time segments) | No | No | No | No | No | No | Yes(Manhattan) | Ranking | Precision and Recall | No(BPR, NMF, ESSVM, LBIMC, LSTM) | No | Yes(BPR, CF) | No | No | Check-ins | No | No | Yes | No | No | No | Yes | Prev-No filtering | No | Foursquare: They refer to Vrer: context- based venue recommendation using embedded space ranking svm in location-based social network 2017 and Noise-tolerance matrix completion for location recommendation 2017 | ||||||||||||||||||||||
214 | 2017 | TSG | A Point of Interest Recommendation Approach by Fusing Geographical and Reputation Influence on Location Based Social Networks | Zeng, J., Li, F., Wen, J., Zhou, W. | https://link.springer.com/chapter/10.1007%2F978-3-030-00916-8_22 | CollaborateCom | Conference | 1 | 0 | \cite{DBLP:conf/colcom/ZengLWZ17} | Yes | No | No(very simple) | No | Yes(PageRank) | No | No | Yes | No | Yes | Yes(Not sure if they are friends, but they build a graph) | No | No | No | No | Yes | No | No | No | No | Yes(70% for training 30% for test) | No | No | No | Yes(removed users and POIs with less than 5 Check-ins) | No | Ranking | Precision and Recall | No(GM-FCF, userCF, FCF, USG) | No | Yes(UCF) | Yes(USG) | No | Check-ins | No | No | No | No | Brightkite | No(They refer to Friendship and mobility: user movement in location- based social networks 2011) | Not complete (only Check-ins) | None | No | Birghtkite: They refer to Friendship and mobility: user movement in location- based social networks 2011 | ||||||||||||||||||||||
215 | 2017 | SSLR | Personalized location recommendation for location-based social networks | Xu, Q., Wang, J., Xiao, B. | https://ieeexplore.ieee.org/document/8330459 | ICCC | Conference | 1 | 2 | \cite{DBLP:conf/iccchina/XuWX17} | No | No | Yes | No | Yes(random walk model) | Yes(geographical + user interaction with temporal information) | No | Yes | No | Yes | Yes | No | No | Yes | No | Yes | No | No | Yes(80% for training 20% to test) | No | No | No | No | No | No | No | Ranking | Precision and Recall | No(USG, iGSLR, ASMF, PMF) | No | No(PMF) | Yes(USG, IGSLR) | No | Check-ins | No | Yes | No | No | Brightkite | No | Yes | Prev-No filtering | No | Brighkite and Gowalla: no further information | ||||||||||||||||||||||
216 | 2017 | LTMM + LAS (LST is the combination of both) | Point-of-interest recommendations: Capturing the geographical influence from local trajectories | Shi, Y., Jiang, W. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8367399 | ISPA/IUCC | Conference | 1 | 3 | \cite{DBLP:conf/ispa/ShiJ17} | Yes(part of the byhrid approach) | No | No | No | No | Yes(combine trajectory model + local activity similarity) | No | No | No | Yes | No | No | No | Yes | No | Yes | No | No | No | No | No | Yes(for each user 30% of the unvisited POIs to test) | No | No | Yes(only users that have visited more than 30 locations) | No | Ranking | Precision and Recall | No(U, PD,LTMM, LAS, LST) | No | Yes(UB) | Yes(PD, power law) | No | POIs | No | No | No | No | No information of the sources | No | Yes | Post | No | ??? | ||||||||||||||||||||||
217 | 2017 | ---No-Acronym-- | A personalized recommendation framework with user trajectory analysis applied in Location-Based Social Network (LBSN) | Xing, L.G., Abiodun, I.A., Khuen, C.W., Boon, T.T. | https://ieeexplore.ieee.org/document/8324177 | ICETSS | Conference | ??? | 1 | 4 | \cite{Xing2017} | Not sure if it is on the scope. Although it seems to recommend POIs, it uses trajectories. --> the task fits the survey, so we should consider it. Move this paper to 2017 | No | No | No | No | Yes(part of the hybrid approach. HITS) | No(Link analysis + KDI. They say it is hybrid but i do see how it is integrated) | No | Yes | No | Yes(clustering and creating regions) | No | No | No | Yes(Trajectory) | Yes(I think yes because apart from the sequences they say "Also, time of visit is equally noted or captured to determine the recency (how recent) or when last the user visited the object of belief") | Yes | No | No | No | No | No | Yes(no info but we will assume random Fix) | No | No | No | No | Ranking | Precision, Recall | Yes(Random, Rank by count, rank by frequency, link analysis, CF) | Yes(Pop, random) | Yes(UB) | No | No | Check-ins | No(They discuss about it) | No | No | No | UniCAT | No | Not complete (only trajectories and stay points) | None | No | Unicat: no further details | ||||||||||||||||||||
218 | 2017 | ---No-Acronym-- | Personalized Point of Interest Recommendation Using Check-In History and Friend's Interest | Erande, D.J., Chaugule, A. | https://ieeexplore.ieee.org/document/8463698 | ICCUBEA | Conference | ??? | 1 | 0 | \cite{Erande2017} | Not sure if it is on the scope, they talk about route recommendation. --> one module is called travel route but the task remains to be POI recommendation. Pablo: I still think it is not on th scope. Very few information about the POI recommendation approach. Nevertheless, we will consier Move this paper to 2017 | Yes | No | No | No | No | Yes(In the mathematical evolution, the prediction is a hybrid approach but im not sure if applies for POI recommendation) | No | No | No | No(I think it is only used for evaluation) | Yes | Yes(tags) | No | No | No | Yes | No | No | No | No | No | No | No | No | No | No | Ranking | Precision, Recall | No(CF) | No | Yes(CF) | No | No | No information | No | No | Yes | No | No | No | No | None | No | Foursquare: no further details | ||||||||||||||||||||
219 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
220 | 2018 | ---No-Acronym-- | Time-slot-based point of interest recommendation on location-based social network | Zeng, J., Li, Y., Li, F., He, X., Wen, J. | ??? | International Journal of Internet Manufacturing and Services | Journal | ??? | 1 | 1 | IGNORE | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
221 | 2018 | RecNet | RecNet: a deep neural network for personalized POI recommendation in location-based social networks | Ding, R., Chen, Z. | International Journal of Geographical Information Science | Journal | 1 | 26 | IGNORE | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
222 | 2018 | ??? | Combining Geographical and Social Influences with Deep Learning for Personalized Point-of-Interest Recommendation | Guo, J., Zhang, W., Fan, W., Li, W. | Journal of Management Information Systems | Journal | 1 | 14 | IGNORE | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
223 | 2018 | ??? | A study of neighbour selection strategies for POI recommendation in LBSNs | Rios, C., Schiaffino, S., Godoy, D. | https://journals.sagepub.com/doi/abs/10.1177/0165551518761000 | Journal of Information Science | Journal | 1,3 | 2 | IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
224 | 2018 | ---No-Acronym-- | Location recommendation with social media data | Bothorel, C., Lathia, N., Picot-Clemente, R., Noulas, A. | https://link.springer.com/chapter/10.1007%2F978-3-319-90092-6_16 | Lecture Notes in Computer Science | Chapter | 1 | 24 | Do not propose any approach. Not on the scope. IGNORE --> it is like a survey | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
225 | 2018 | ---No-Acronym-- | User Modeling for Point-of-Interest Recommendations in Location-Based Social Networks: The State of the Art | Liu, S. | https://www.hindawi.com/journals/misy/2018/7807461/ | Mobile Information Systems | Journal | 1 | 10 | Do not propose any approach. Not on the scope. It is some kind of survey. IGNORE --> it is a survey | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
226 | 2018 | ---No-Acronym-- | An automatic user grouping model for a group recommender system in location-based social networks | Khazaei, E., Alimohammadi, A. | https://www.mdpi.com/2220-9964/7/2/67 | ISPRS | Journal | 1 | 7 | Group recommendation. IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
227 | 2018 | ---No-Acronym-- | Objectives and state-of-the-Art of location-Based social network recommender systems | Ding, Z., Li, X., Jiang, C., Zhou, M. | https://dl.acm.org/citation.cfm?id=3154526 | ACM Computing Surveys | Journal | 1,3 | 20 | Survey. IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
228 | 2018 | FDPL | Friend recommendation in location-based social networks via deep pairwise learning | Rafailidis, D., Crestani, F. | https://ieeexplore.ieee.org/document/8508362 | ASONAM | Conference | 1 | 5 | Friend recommendation. IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
229 | 2018 | ---No-Acronym-- | Privacy-Preserving POI Recommendation Using Nonnegative Matrix Factorization | Wang, X., Yang, H., Lim, K. | https://ieeexplore.ieee.org/document/8511836 | PAC | Conference | 1 | 2 | Not sure if it is on the scope. It transform private Check-ins for into groups and it only 2 pages with no results. IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
230 | 2018 | ---No-Acronym-- | Time-aware location sequence recommendation for cold-start mobile users | Shen, T., Chen, H., Ku, W.-S. | https://dl.acm.org/citation.cfm?doid=3274895.3274958 | SIGSPATIAL | Conference | 1 | 3 | No results provided. IGNORE --> it proposes a method to predict a sequence of categories, but the task is still POI recommendation, although it has no experiments | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
231 | 2018 | ---No-Acronym-- | Preference Aware Travel Route Recommendation with Temporal Influence | Madhuri Debnath, Praveen Kumar Tripathi, Ashis Kumer Biswas, Ramez Elmasri | https://dl.acm.org/doi/abs/10.1145/3282825.3282829 | SIGSPATIAL | Conference | 3 | 0 | Travel recommendation,. Tutorial. IGNORE --> it is not a tutorial but the task is out of scope <-- it should be considered in our UMUAI_SI paper | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
232 | 2018 | UI-GF | Item-driven group formation | Daniel Valcarce, Igo Ramalho Brilhante, Jose A. F. de Macedo, Franco Maria Nardini, Raffaele Perego, Chiara Renso | https://www.sciencedirect.com/science/article/pii/S246869641830096X | Online Social Networks and Media | Journal | 2 | 0 | Group recommendation. IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
233 | 2018 | PK-Boosting | Personalized context-aware point of interest recommendation | Aliannejadi, M., Crestani, F. | https://dl.acm.org/citation.cfm?doid=3211967.3231933 | TOIS | Journal | 1 | 34 | \cite{DBLP:journals/tois/AliannejadiC18} | I think is tag prediction. --> one of the tasks is tag prediction, but the main focus and several of the experiments are about POI recommendation. In section 4 they discuss about POI recommendation. IGNORE because there are no Check-ins! | No | No | Yes | No | No | Yes | Expectation Maximization | No | No | No | No | Yes | Yes | No | No | No | No | No | No | No | No | In both TREC-CS 2015 and TREC-CS 2016 datasets, for each user u,the data, S(u), is split into two sets: a number of locations visited in one or two cities con- stitute the training set and a number of locations in a new city constitute the test set | No(already filtered) | No | Ranking | Precision, NDCG and MRR | No(LinearCatRev, GeoSoca, nDTF) | No | No | Yes(Geosoca) | No | No information. I would say Check-ins | No | No | No | No | TREC | No(They refer to Overview of the TREC 2015 contextual suggestion track 2015 and Overview of the TREC 2016 contextual suggestion track 2016) | Not complete (Users and instances) | None | No | TREC: no further details | |||||||||||||||||||||||
234 | 2018 | CLoSe | Close: Contextualized location sequence recommender | Baral, R., Iyengar, S.S., Li, T., Balakrishnan, N. | https://dl.acm.org/citation.cfm?doid=3240323.3240410 | RecSys | Conference | 1,3 | 9 | \cite{DBLP:conf/recsys/BaralI0018} | It recommends trajectories, or at least it uses pair F1 measure, taking into account the order. IGNORE | No | No | No | Yes | No | No | No | No | No(neural network, not embedding) | Yes | No | Yes | No | Yes(uses a sequential vector) | Yes | No | No | No | No | No | No | Yes(For each user, 10 most frequency checked POIs were taken as starting POIs and 10 sequences per starting point were generated) | Yes(users with less than 10 valid sequences ignored) | No | Ranking | Precision Pair, Recall Pair, F1 Pair, Diversity, Displacement | Yes(Popularity, Apriori, Markov, HITS, RNN, LSTM) | Yes(Pop) | No | No | No | POIs | No | Yes | No | No | Weeplaces | Yes(http://www.yongliu.org/datasets/ for Weeplaces and they also refer to Personalized point-of-interest recommendation by mining users’ preference transition 2013) | Yes | Prev | No | Weeplaces and Gowalla: Personalized point-of-interest recommendation by mining users’ preference transition 2013 | |||||||||||||||||||||||
235 | 2018 | Maybe QSim ---No-Acronym-- | A collaborative recommendation system for location based social networks | Belkhir, A., Bouyakoub, F.M., Boubenia, M. | https://ieeexplore.ieee.org/document/8379012 | ISPS | Conference | ??? | 1 | 0 | IGNORE. No experiments provided | Yes | No | No | No | No | No | No | No | No | Yes | No | Yes(attributes of the user and items) | No | No | No | No Information | No Information | No Information | No Information | No Information | No Information | No | No | None | None | No | No | No | No | No | No information | No(they discuss about it) | No | No | No | No | No | No | None | No | ??? | ||||||||||||||||||||||||
236 | 2018 | LBPR | Next point-of-interest recommendation via a category-aware Listwise Bayesian Personalized Ranking | He, J., Li, X., Liao, L. | Journal Computation Science | Journal | 1,2 | 7 | \cite{DBLP:journals/jocs/HeLL18} | Next POI recommendation. IGNORE. Repeated in 2017 \cite{DBLP:conf/ijcai/HeLL17} | No | Yes(MF optimized by an approach similar to BPR) | Yes(BPR and transition probabilities) | No | No | No | Stochastic Gradient Descent | No | No | Yes(distance) | No | Yes(category) | No | Yes(Transition) | No | No | Yes(for every user 80% of her recent Check-ins to train, rest to test) | No | No | No | No | Yes(removed users with less than 20 Check-ins) | Yes(New York and Los Angeles) | Ranking | Precision and New Precision (for new POIs) | No(FPMC-LR, PRME-G, LBP, LBPR-M-First, LBPR-N-First, LBPR-M-Second, LBPR-N-Second) | No | No | No | No | Check-ins (but NON overlapping) | No | No | Yes | No | No | No | Yes | Post | No | Foursquare: no further details | |||||||||||||||||||||||||
237 | 2018 | ATTF | Aggregated temporal tensor factorization model for point-of-interest recommendation | Zhao, S., Lyu, M.R., King, I. | https://link.springer.com/article/10.1007%2Fs11063-017-9681-8 | Neural Processing Letters | Journal | 1 | 12 | \cite{DBLP:journals/npl/ZhaoKL18} | Repeated in 2016. IGNORE. The same as \cite{DBLP:conf/iconip/ZhaoLK16} | No | Yes | No | No | No | No | No | No | No | No | No | No | No | No | Yes | No | No | No | Yes(80%-20% of the Check-ins of each user) | No | No | Yes(removed POIs checked-in by less than 5 users and removed users with less than 10 Check-ins) | No | Ranking | Precision, Recall, F score | No(ERMF, BPR-MF, LRT. FPMC-LR) | No | Yes(BPRMF, WRMF) | No | No | Check-ins | No | Yes | Yes | No | No | No(They refer to paper: gscorr: modeling geo-social correlations for new check-ins on location-based social networks (2012) for foursquare and Capturing geographical influence in poi recommendations. In: Interna- tional conference on neural information processing for Gowalla (2013)) | Yes | Post | No | Foursquare and Gowalla: gscorr: modeling geo-social correlations for new check-ins on location-based social networks 2012 for Foursquare Aggregated temporal tensor factorization for point-of-interest recommendation 2016 Gowalla | ||||||||||||||||||||||||
238 | 2018 | WPOI | Investigating the utility of the weather context for point of interest recommendations | Trattner, C., Oberegger, A., Marinho, L., Parra, D. | https://link.springer.com/article/10.1007/s40558-017-0100-9 | Information Technology & Tourism | Journal | 1 | 9 | \cite{DBLP:journals/jitt/TrattnerOMP18} | Repeated in 2016. IGNORE. The same as \cite{DBLP:conf/recsys/TrattnerOEPM16} | No | Yes | No | No | No | No | Stochastic Gradient Descent | No | No | Yes | No | Yes(Weather) | No | No | Yes(Depending on RankGeo) | No | Yes(70% training, 20% test and 10% validation) | No | No | No | No | Yes(users with 20 interactions and POIs visited 2 times) | Yes(Minneapolis, Boston, MIami and Honolulu) | Ranking | NDCG | No(Rank-GeoFM and variations) | No | No | Yes(RankGeo-FM) | Yes | Check-ins | No | No | Yes | No | No | No(They refer to Participatory cultural mapping based on collective behavior data in location-based social networks 2016) | Yes | Prev and Post | Yes(https://github.com/aoberegg/WPOI) | Yes | No | Yes | Foursquare: They refer to Participatory cultural mapping based on collective behavior data in location-based social networks 2016 | |||||||||||||||||||||
239 | 2018 | ICCF | Scalable Content-Aware Collaborative Filtering for Location Recommendation | Lian, D., Ge, Y., Zhang, F., Yuan, N.J., Xie, X., Zhou, T., Rui, Y. | https://ieeexplore.ieee.org/document/8246576 | IEEE Transactions on Knowledge and Data Engineering | Journal | 1 | 39 | \cite{DBLP:journals/tkde/LianGZY0ZR18} | Repeated in 2015. IGNORE. The same as \cite{DBLP:conf/icdm/LianGZYXZR15} | No | Yes | No | No | No | No | Alternate Least squares | No | No | No | No | Yes | Yes | No | No | No | No | No | No | Yes (5-Fold cross validation at user level, for outMatrix) | Yes (5-Fold cross validation at user level, for inMatrix) | Yes(only users that have visited at least 10 POIs and POIs visited by at least 10 users) | No | Ranking | Precision and Recall | No(LibFM,GRMF,LightFM,GeoMF) | No | No | Yes(GeoMF) | No(5-fold validation) | Check-ins for in matrix and users for out matrix | No | No | No | No | Jiepang (chinese LBSN similar to Foursquare) | No | Yes | Post | No(They refer to other sources, from libraries and I think that for baselines) | Jiepang: no further details | ||||||||||||||||||||||||
240 | 2018 | TM-PFM | Exploiting human mobility patterns for point-of-interest recommendation | Yao, Z. | https://dl.acm.org/citation.cfm?doid=3159652.3170459 | WSDM | Conference | 1 | 10 | \cite{DBLP:conf/wsdm/Yao18} | Only 2 pages. Doctoral presentation. IGNORE. Repeated in 2016 \cite{DBLP:conf/icdm/YaoFLLX16} | No | Yes(Poisson factor model) | Yes(Poisson factor model) | No | No | No | No | No | No | No | No | Yes | No | No | Yes | No Information | No Information | No Information | No Information | No Information | No Information | No | Yes(New York) | Ranking | F-measure | No(PMF, BPTF-Voting, BPTF-Sum, NMF, LRT-Voting, LRT-Sum) | No | No(PMF) | No | No | No information but NOT visited in the training set | No | No | Yes | No | No | No | No | None | No | Foursquare: no further information | ||||||||||||||||||||||||
241 | 2018 | GeoMF++ | GeoMF++: Scalable location recommendation via joint geographical modeling and matrix factorization | Lian, D., Zheng, K., Ge, Y., Cao, L., Chen, E., Xie, X. | https://dl.acm.org/citation.cfm?doid=3146384.3182166 | TOIS | Journal | 1,3 | 52 | \cite{DBLP:journals/tois/LianZGCCX18} | Repeated in 2014. IGNORE. The same as \cite{DBLP:conf/kdd/LianZXSCR14} | No | Yes | No(They say the apply kernels) | No | No | No | Alternate Least squares | No | No | Yes | No | No | No | No | No | No | No | No | No | No | Yes (5 cross-validation at user level) | Yes(users and pois with less than 10 visits, removed) | Yes(For Jiepang, they use Beijing) | Ranking | Recall and NDCG | No(GWMF, WRMF, IRenMF, UCF, HPF, BPRMF) | No | Yes(BPR) | Yes(IRenMF) | No(Cross-validation) | Check-ins, POIs | Yes(for new locations) | Yes | No | No | Jiepang (chinese LBSN similar to Foursquare) | Refer to other paper | Yes | Post | Yes(https://github.com/DefuLian/recsys.git) But in that URL I do not find anything about GeoMF++ | Yes | Yes | Yes | Gowalla: They refer to Friendship and mobility: user movement in location-based social networks 2011 | |||||||||||||||||||||
242 | 2018 | ---No-Acronym-- | Fine-Gained Location Recommendation Based on User Textual Reviews in LBSNs | Chen, Y., Zheng, Z., Sun, L., Chen, D., Guo, M. | https://link.springer.com/chapter/10.1007%2F978-3-030-15093-8_14 | GPC | Conference | 1 | 0 | \cite{DBLP:conf/gpc/ChenZ00G18} | Yes | No | No | No | No | Yes(KNN + features) | Yes(Learning to rank + elo) | No | No | No | No | No | Yes(sentiment and categories) | Yes | No | No | Yes | No | No | No | No | Yes(80% of the data for training, rest to test) | No | No | No | Yes(remove URLs and no chinese words) | No | Ranking | Precision and Recall | No(BiasMF, SVD++, ORec, AspectRec, TriRank, TopicMF) | No | Yes(SVD++, Biased MF) | Yes(ORec) | No | POIs(Equivalent) | Yes | No | No | No | DianPing | No(They refer to this page http://www.dianping.com/ but i dont think we can obtain the dataset there) | Yes (but they are splitted in different types of venues) | Post | No | Diaping: they only provide Dianping url, nothing more | |||||||||||||||||||||
243 | 2018 | ---No-Acronym-- | A self-adaptive point-of-interest recommendation algorithm based on a multi-order Markov model | Liu, S., Wang, L. | Future Generation Computer Systems | Journal | 1,2 | 12 | \cite{DBLP:journals/fgcs/LiuW18} | No | No | Yes(Markov model) | No | No | Yes(Eq 14) | No | No | No | Yes | No | No | No | Yes | Yes | Yes | No | No | No | No | Yes, but it is 60% training, 20% validation and 20% test | No | No | No | No | No | Ranking | F-measure(also define Precision and Recall) | No(LORE, RankGeoFM, HSMM) | No | No | Yes(RankGeo-FM, LORE) | Yes(20%) | Check-ins | No | Yes | No | No | Brightkite | No(They refer to Friendship and mobility: user movement in location-based social networks 2011) | Yes | Prev-No filtering | No | Brightkite and Gowalla: They refer to Friendship and mobility: user movement in location-based social networks 2011 | |||||||||||||||||||||||
244 | 2018 | PCRM | PCRM: Increasing POI recommendation accuracy in location-based social networks | Liu, L., Li, W., Wang, L., Jia, H. | http://itiis.org/digital-library/manuscript/2173 | Transactions on Internet and Information Systems | Journal | 1 | 0 | \cite{DBLP:journals/itiis/LiuLWJ18} | Activity Recommendation | No | Yes(They say it can be interpreted of Extension of LDA) | Yes(Similar to LDA) | No | No | No | No | No | No | No(Im not sure if it is specifically modeled, although it is used for the regions) | No | Yes | No | No | No | Yes | No(although they claim to use a online recommendation) | No | No | No | Yes(90%training, rest to test) | No | No | No | Yes(but no data provided) | No | Ranking | Accuracy | No(PCM, PRM, IKNN, CKNN, USG) | No | Yes(IKNN) | Yes(USG) | No | Check-ins | No | No | No | No | Evenbrite | No | Yes | Post | No | Evenbrite: no further details | |||||||||||||||||||||
245 | 2018 | STSCR | STSCR: Exploring spatial-temporal sequential influence and social information for location recommendation | Gao, R., Li, J., Li, X., Song, C., Chang, J., Liu, D., Wang, C. | https://www.sciencedirect.com/science/article/pii/S0925231218308762 | Neurocomputing | Journal | 1,2 | 8 | \cite{DBLP:journals/ijon/GaoLLSCLW18} | No | Yes(tensor factorization) | Yes(BPR and maximizing posterior) | No | No | No | Stochastic Gradient Descent | No | No | No | Yes | No | No | Yes | Yes | Yes | No | No | No | No | No | No | Yes (5-Fold cross validation at the dataset level) | No | Yes(remove users and POIs with less than 10 Check-ins) | No(specifically stated all over the world) | Ranking | Precision, Recall, MAP, NDCG | No(BPRMF, USG, FPMC+LR, LTSCR, LRT, SPRE, LORE) | No | Yes(BPR) | Yes(USG, LORE) | No(5 fold-cross validation) | Check-ins | No | Yes | Yes | No | No | No(They refer to gscorr: modeling geo-social correlations for new check- -ins on location-based social networks 2012 for Foursquare and Friendship and mobility: user movement in location-based social networks 2011 for Gowalla) | Yes | Post | No | Foursquare and Gowalla: gscorr: modeling geo-social correlations for new check- -ins on location-based social networks 2012 for Foursquare and Friendship and mobility: user movement in location-based social networks 2011 for Gowalla | ||||||||||||||||||||||
246 | 2018 | TransTL | Time and Location Aware Points of Interest Recommendation in Location-Based Social Networks | Qian, T.-Y., Liu, B., Hong, L., You, Z.-N. | https://link.springer.com/article/10.1007%2Fs11390-018-1883-7 | Journal of Computer Science and Technology | Journal | 1 | 7 | \cite{DBLP:journals/jcst/QianLHY18} | No | Yes(Graph embedding) | No | No | No | No | Gradient descent | No | Yes(graph embedding) | Yes | No | No | No | No | Yes | Yes | No | No | No | Yes(70% training, 10% validation and 20% test for each user, ordered by timestamps) | No | No | No | No | No | No | Ranking | Accuracy, Recall | No(GE, TransRec, TransTL-E, TransTL-H) | No | No | Yes(Ge) | Yes(10% for each user for validation) | Check-ins | Yes(cold start POIs, with less than 5 interactions) | Yes | Yes | No | No | Yes(https://sites.google.com/site/dbhongzhi/) | Yes | Prev-No filtering | No | Foursquare and Gowalla: They use this dataset https:/sites.google.com/site/dbhongzhi, provided by Learning graph-based POI embedding for location-based recommendation 2016 | ||||||||||||||||||||||
247 | 2018 | ALGeoSPF | Algeospf: A hierarchical factorization model for poi recommendation | Griesner, J.-B., Abdessalem, T., Naacke, H., Dosne, P. | https://ieeexplore.ieee.org/document/8508249 | ASONAM | Conference | 1 | 1 | \cite{DBLP:conf/asunam/GriesnerAND18} | No | Yes(they say it is based on Poisson MF, but I do not see the formulation) | Yes(they say it is based on Poisson MF, but I do not see the formulation) | No | No | No | No | Yes | No | Yes | No(not clear if the social component exploit friends of the user or just neighbours) | No | No | Yes(Transition) | No | Yes | No | No | No Information | No Information | No Information | No Information | No Information | No Information | No | No | Ranking | Recall and NDCG | No(PF, BPR, SLIM, PMF, NMF) | No | Yes(BPR) | No | No | No information | No | Yes | Yes | No | YFCC | No(They refer to Time-aware point-of-interest recommendation 2013 for Foursquare and Friendship and mobility 2011 for Gowalla) | Yes | Prev-No filtering | No | Gowalla, Foursquare and YFCC: They refer to Time-aware point-of-interest recommendation 2013 for Foursquare and Friendship and mobility 2011 for Gowalla | ||||||||||||||||||||||
248 | 2018 | SAE-NAD | Point-of-interest recommendation: Exploiting self-attentive autoencoders with neighbor-aware influence | Ma, C., Wang, Q., Zhang, Y., Liu, X. | https://dl.acm.org/citation.cfm?doid=3269206.3271733 | CIKM | Conference | 1 | 36 | \cite{DBLP:conf/cikm/MaZWL18} | Yes | No | No | No | Yes(Encoder/Decoder) | No | No | No | No | No(It is a encoder/decoder and althugh it uses embeddings it is not graph nor word embedding) | Yes | No | No | No | No | No | Yes | No | No | No | No | No | Yes(for each user, 20% of visited locations to test) | No | No | Yes(For Gowalla, remove users and POIs with less than 20 visits, For Foursqaure and Yelp, remove users and POIs with less than 10 visits) | Yes(For foursqaure, USA, rest no info) | Ranking | Precision, Recall and MAP | No(WRMF, BPRMF, MGMMF, IrenMF, RankGeoFM, PACE, DeepAE) | No | Yes(BPR, MF) | Yes(IrenMF, MGMMF) | No | POIs | No | Yes | Yes | Yes | No | No(They refer to Friendship and mobility: user movement in location-based social networks 2011 for Gowalla and An Experimen- tal Evaluation of Point-of-interest Recommendation in Location-based Social Networks 2017 for Foursquare and Yelp) | Yes | Post | Yes(https://github.com/allenjack/SAE-NAD) | Yes | No | Yes | Gowalla, Foursquare and Yelp: Friendship and mobility: user movement in location-based social networks 2011 for Gowalla, An Experimen- tal Evaluation of Point-of-interest Recommendation in Location-based Social Networks 2017 for Foursquare and Yelp | ||||||||||||||||||
249 | 2018 | LTSR | Location-Time-Sociality Aware Personalized Tourist Attraction Recommendation in LBSN | Zhu, Z., Cao, J., Weng, C. | https://ieeexplore.ieee.org/document/8465179 | CSCWD | Conference | 1 | 4 | \cite{DBLP:conf/cscwd/ZhuCW18} | Tourist attraction | Yes(part of the hybrid approach) | Yes(LDA) | Yes(LDA) | No | No | Yes(social + location distance + time-aware pop) | No | Yes(Density Based Spatial Clustering of Applica- tions with Noise (DBSCAN)) | No | Yes | Yes | No | No | No | Yes | Yes | No | No | Yes(10% of the most recent Check-ins to test, rest to train) | No | No | No | No | No | Yes(remove Check-ins whose categories are not tourist attractions) | Yes(New York and San Francisco) | Ranking | Precision and Recall | No(MV, PB, TACF, LDA, LLDA, LLDA_P, LLDA_D, LTSR) | Yes(MV is popularity) | No | Yes(LLDA_D uses distance) | No | Check-ins | No | No | Yes | No | No | No | Yes | Post | No | Foursquare: no further details | |||||||||||||||||||||
250 | 2018 | GR-DELM | A novel recommendation system in location-based social networks using distributed ELM | Zhao, X., Ma, Z., Zhang, Z. | https://link.springer.com/article/10.1007%2Fs12293-017-0227-4 | Memetic Computing | Journal | 1 | 7 | \cite{DBLP:journals/memetic/ZhaoMZ18} | Yes | Yes | No | No | Yes(katz method) | No(although in algorithm 1 in line 8 can be considered as hybrid, it is in the first part of the algorithm, so no) | No | No | No | Yes | Yes | No | No | No | No | Yes | No | No | No | No | Yes(80% training and 20% test random) | No | No | No | No | No | Ranking | Precision, Recall, F1-measure | No(HGSMR, RFR, UCFR, EPR) | No | Yes(UCF) | No | No | Check-ins | No | Yes | No | No | Brightkite | No | Not complete (number of POIs not stated) | None | No | Gowalla and Brightkite: no further details | ||||||||||||||||||||||
251 | 2018 | ReGS | Points-of-interest recommendation based on convolution matrix factorization | Xing, S., Liu, F., Zhao, X., Li, T. | https://link.springer.com/article/10.1007%2Fs10489-017-1103-0 | Applied Inteligence | Journal | 1 | 15 | \cite{DBLP:journals/apin/XingLZL18} | Yes(Compute similarities between users and friends. I would vote for Yes) | Yes(Most important) | Yes(PMF) | Yes | No | No | Gradient descent | No | No | Yes | Yes(friends) | No | Yes(reviews, semantics) | No | No | Yes | No | No | Yes(Half of the Check-ins with earlier timestamps for training and rest for test) | No | No | No | No | No | No | Yes(New York and Los Angeles) | Error/Ranking | RMSE, Precision and Recall | No(LCARS, CoRe, GeoMF, DRW, NCPD) | No | No | Yes(GeoMF) | No | Check-ins | No(although they discuss about it) | No | Yes | No | No | No(They say they use the same crawling strategy of Exploring social-historical ties on location-based social networks 2012) | Yes | Prev-No filtering | No | Foursquare: They say they use the same crawling strategy of Exploring social-historical ties on location-based social networks 2012 | ||||||||||||||||||||||
252 | 2018 | LocRec | LocRec: Rule-based successive location recommendation in LBSN | Amirat, H., Benslimane, A., Fournier-Viger, P., Lagraa, N. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8422183 | ICC | Conference | 1 | 1 | \cite{DBLP:conf/icc/AmiratBFL18} | Successive POI recommendation | No | No | No | No | No | No | Yes. Rule-based(rule mining and rule matching for recommendation) | No | No | No | No | Yes | No | No | Yes | Yes | Yes | No | No | No | No | No | Yes("We first, extract the daily location sequences of each user to construct his location history. Once location histories of all users are extracted, a random division of the locations histories, according to a parameter called Training ratio, into training and testing sets is performed. Training set is used to generate recommendation rules while the testing set is used for testing") | No | No | No | No | Ranking | Accuracy and Coverage | No(SSR-Rec, SP-Rec) | No | No | No | No | Check-ins | No | Yes | No | No | No | Yes(https://snap.stanford.edu/data/) | Not complete (Only users) | None | No | Gowalla: https://snap.stanford.edu/data/ | ||||||||||||||||||||
253 | 2018 | ULE | ULE: Learning user and location embeddings for POI recommendation | Wang, H., Ouyang, W., Shen, H., Cheng, X. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8411844 | DSC | Conference | 1 | 4 | \cite{DBLP:conf/dsc/WangOSC18} | Yes | Yes(They discuss about latent factors) | Yes | No | No | Yes(combination of three models) | adopt asynchronous stochastic gradient descent (ASGD) | No | Yes(embedding vectors) | Yes | Yes | No | No | No | No | Yes | No | No | No | No | No | Yes(70% training, 15% validation, 15% test for each user for POIs, not Check-ins) | No | No | No | No | Ranking | Precision and Recall | No(UCF, MF-Count, MF-O1, MF-Weight, MF-COnfidence, PFM, MF-BPR, UCF+GS/G/S, PMF+MGM, GeoMF, ASMF/ARMF) | No | Yes(UCF) | Yes(GeoMF) | Yes(15%) | POIs | No | Yes | Yes | No | No | No | Not complete (number of POIs not stated) | None | No | Foursquare and Gowalla: no further details | ||||||||||||||||||||||
254 | 2018 | Maybe UZT --No-Acronym--- | POI recommendation of location-based social networks using tensor factorization | Liao, G., Jiang, S., Zhou, Z., Wan, C., Liu, X. | https://ieeexplore.ieee.org/document/8411268 | MDM | Conference | 1 | 11 | \cite{DBLP:conf/mdm/LiaoJZWL18} | No | Yes(tensor factorization + LDA) | Yes(LDA) | No | No | No | higher order singular value decomposition (HOSVD) | No | No | No | No | No | Yes(comments) | No | Yes | Yes | No | No | No Information | No Information | No Information | No Information | No Information | No Information | No | No | Ranking | Precision, Recall, MAP | No(TLA, CMF, ULT) | No | No | No | No | No information | No | No | No | No | WW | No | Yes | Prev-No filtering | No | WW: no further details | ||||||||||||||||||||||
255 | 2018 | ReEl and DAP | ReEL: Review aware explanation of location recommendation | Baral, R., Zhu, X.L., Iyengar, S.S., Li, T. | https://dl.acm.org/citation.cfm?id=3209219.3209237 | UMAP | Conference | 1,3 | 17 | \cite{DBLP:conf/um/BaralZIL18} | No | Yes | No | Yes | No | No | No | No | No(They use embeddings but in DNN) | Yes | No | Yes(categorical) | Yes(textual) | No | No | Yes | No | No(There is a case study but it is not a per user study) | No | No | No | No | Yes | No | No | No | Ranking | Precision, Recall, F score | No(UCF, ICF, PPR, ORec, LDA, Embedding, DeepConn, DAP, AGSG) | No | Yes(UCF, ICF) | Yes(AGSG) | No(5-fold validation) | POIs(Equivalent) | No | No | No | Yes | Yelp, TripAdvisor, Airbnb | Yes(https://www.yelp.com/dataset_challenge for Yelp, http://insideairbnb.com/get-the-data.html for Airbnb and Latent aspect rating analysis without aspect keyword supervision 2011 for tripdavisor) | Yes | Prev-No filtering | No | Yelp, TripAdvisor, Airbnb: https://www.yelp.com/dataset_challenge for Yelp Latent aspect rating analysis without aspect keyword supervision 2011 for TripAdvisor http://insideairbnb.com/get-the-data.html for aibnbn | ||||||||||||||||||||||
256 | 2018 | GSBPR | Exploiting geo-social correlations to improve pairwise ranking for point-of-interest recommendation | Gao, R., Li, J., Du, B., Li, X., Chang, J., Song, C., Liu, D. | https://ieeexplore.ieee.org/document/8424613 | China communications | Journal | ??? | 1 | 6 | \cite{Gao2018} | No | Yes(They claim to use MF) | Yes(distributions, BPR etc) | No | No | No | Stochastic Gradient descent, BPR | No | No | Yes | Yes | No | No | No | No | Yes | No | No | No | No | No | No | Yes (Random 80-20 repeated 5 times) | No | No | No | Ranking | Precision, Recall, MAP, NDCG | No(USG, IRenMF, MGM, BPR-MF, MBPR, GBPR, SPRE) | No | Yes(BPRMF) | Yes(IrenMF, MGM) | No(5-fold validation) | Check-ins(I think it is by Check-ins) | No | No | Yes | Yes | No | No(They refer to gSCorr: Modeling Geo-social Correlations for New Check-ins on Location-based Social Networks 2012 For Foursquare and GeoSoCa: Exploiting Geo- graphical, Social and Categorical Correlations for Point-of-interest Recommendations 2015 for Yelp) | Yes | Prev-No filtering | No | Foursquare and Yelp: gSCorr: Modeling Geo-social Correlations for New Check-ins on Location-based Social Networks 2012 for Foursquare GeoSoCa: Exploiting Geo- graphical, Social and Categorical Correlations for Point-of-interest Recommendations for Yelp | |||||||||||||||||||||
257 | 2018 | BTC + variants | Cross-urban point-of-interest recommendation for non-natives | Xu, T., Ma, Y., Wang, Q. | International Journal of Web Services Research | Journal | 1 | 3 | \cite{DBLP:journals/jwsr/XuMW18} | No | No | No | No | No | No | Yes(Transfer learning model) | No | Yes | No | No | No | Yes | No | No | Yes | Yes | No | No | No | No | No | No | No | No | Yes(For every user, all Check-ins in residence city in train, rest in test. Other, fix) | Yes(from the original 415 citires, they keep 59US cities) | Yes(for some experiments, they use Wasnighton, Los Angeles, New York and CHicago. U.S. cities) | Ranking | Accuracy | No(CF, MF, CRCF) | No | Yes(MF, CF) | No | No | Check-ins | No | No | Yes | No | No | Yes(https://sites.google.com/site/yangdingqi/home/foursquare-dataset) | Yes | Prev and Post (i will indicate post) | No | Foursquare: https://sites.google.com/site/yangdingqi/home/foursquare-dataset) | |||||||||||||||||||||
258 | 2018 | CARA | A contextual attention recurrent architecture for context-aware venue recommendation | Manotumruksa, J., Macdonald, C., Ounis, I. | https://dl.acm.org/citation.cfm?doid=3209978.3210042 | SIGIR | Conference | 1 | 37 | \cite{DBLP:conf/sigir/ManotumruksaMO18} | Yes | No | No | No | Yes | No | No | No | No | No(They use embeddings but in DNN) | Yes | No | No | No | Yes | Yes | Yes | No | No | No | Yes(Most recent rating as test, then select other 100 venues that has not visited before and perform the evaluation. Leave one out methodology) | No | No | No | No | Yes(removed venues with less than 10 Check-ins) | No | Ranking | Hit rate, NDCG | No(MF, BPR, GeoBPR, STELLAR, NeuMF, DRCF, RNN, STGRU, CAGRU, TimeGRU, CGRU, LatentCross) | No | Yes(BPR, MF) | Yes(GeoBPR) | No | Check-ins | Yes(users with less than 10 Check-ins) | No | Yes | Yes | Brightkite | Yes(https://snap.stanford.edu/data/ for Brighkite, https://archive.org/details/201309_foursquare_dataset_umn for Foursquare and https://www.yelp.com/dataset/challenge for Yelp) | Yes | Post | Yes(https://github.com/feay1234/CARA) | Yes | No | No | Foursquare, Brightkite and Yelp: https://snap.stanford.edu/data/ for Brightkite, https://archive.org/details/201309_foursquare_dataset_umn for Foursquare and https://www.yelp.com/dataset_challenge for Yelp | ||||||||||||||||||
259 | 2018 | GeoUMF | A multi-factor influencing POI recommendation model based on matrix factorization | Xu, Y., Li, Y., Yang, W., Zhang, J. | https://ieeexplore.ieee.org/document/8377512 | ICACI | Conference | 1 | 1 | \cite{DBLP:conf/icaci/XuLYZ18} | No | Yes | No | No | No | No | Stochastic Gradient Descent | No | No | Yes | No | No | No | No | No | Yes | No | No | No | No | Yes(30%training 50% validation and 20% test) | No | No | No | No | Yes(Singapore) | Ranking | Precision and Recall | No(RankGeoMF/G, RankGeoFM) | No | No | Yes(RankGeo-FM) | Yes(50% of validation) | Check-ins | No | No | Yes | No | No | No | Yes | Prev-No filtering | No | Foursquare: no more info | ||||||||||||||||||||||
260 | 2018 | ---No-Acronym-- | Research and implementation of POI recommendation system integrating temporal feature | Liu, J., Jiao, X., Jin, Y., Liu, X., Liu, L. | https://ieeexplore.ieee.org/document/8367694 | ICBDA | Conference | ??? | 1 | 2 | \cite{Liu2018} | No | Yes(I think it is a MF with components of CF | No | No | No | No | No | No | No | No | No | No | No | No | Yes | Yes | No | No | No | Yes(85% for the user for train, next 15% for test) | No | No | No | No | No | No | Ranking | Hit rate, accuracy rate | None | No | No | No | No | Check-ins (but at least they say the check-in matrix has aggregate values) | No | Yes | No | No | No | No | Yes | Prev-No filtering | No | Gowalla: crawled from there | |||||||||||||||||||||
261 | 2018 | ---No-Acronym-- | Mining place-time affinity to improve POI recommendation | Wang, J., Bagul, D., Chu, J., Meng, L., Srihari, S. | https://ieeexplore.ieee.org/document/8356834 | ICICT | Conference | ??? | 1 | 0 | \cite{Wang2018} | No | Yes | No | No | No | No | Yes(MF, gradient boosting decision tree) | No | No | No(MF embedding but not anything else) | Yes(distance to the user. I would vote as yes because it not for filtering) | No | Yes | No | No | Yes | Yes | No | No | No | No | No | No | No | No | Yes(It seems they use Tokyo for training and NY for test) | No | Yes(Tokyo, New York) | Ranking | MRR | None(a baseline they program ad-hoc) | No | No | No | No | Check-ins | No | No | Yes | No | No | No(They refer to A survey of collaborative filtering techniques 2009 “Learning the parts of objects by non- negative matrix factorization 1999 and Participatory cultural mapping based on collective behavior data in location-based social networks 2016) | Not complete (only eheckins stated) | None | No | Foursquare: they refer to A survey of collaborative filtering techniques 2009, Learning the parts of objects by non- negative matrix factorization 1999 and Participatory cultural mapping based on collective behavior data in location-based social networks 2016. Stat matches other works but they are not complete | |||||||||||||||||||
262 | 2018 | NBPR | Location regularization-based POI recommendation in location-based social networks | Guo, L., Jiang, H., Wang, X. | https://www.mdpi.com/2078-2489/9/4/85 | Information | Journal | 1 | 6 | \cite{DBLP:journals/information/GuoJW18} | Yes | No | Yes(Because it is a BPR for NN) | No | No | No | BPR, SGD | No | No | Yes | No | No | No | No | No | Yes | No | No | No | No | No | Yes(80% random of each users Check-ins) | No | No | Yes(only users that have visited moren than 3 and POIs visited more than 20 times and for gowallam users need to visit more than 5 and POIs visited more than 30 times) | No | Ranking | Precision and Recall | Yes(Pop, WRMF, GeoMF, BPRMF, IRenMF, WBPR-F) | Yes(POp) | Yes(BPRMF, WRMF) | Yes(GeoMF, IrenMF) | Yes(5 fold validation in the training set) | POIs | No | Yes | No | No | Brightkite | No(They refer to Friendship and mobility: User movement in location-based social networks 2011) | Yes | Post | No | Gowalla and Brightkite: no further details | ||||||||||||||||||||||
263 | 2018 | CUSPG | Point-of-Interest Recommendation Based on Spatial Clustering in LBSN | Su, C., Li, N., Xie, X.-Z. | https://ieeexplore.ieee.org/document/8843194 | ICNISC | Conference | ??? | 1 | 0 | \cite{Chang2018} | Yes(part of the hybrid approach) | No | Yes(part of the hybrid approach, naive bayes) | No | No | Yes(probability with distance + social + UB) | No | Yes(cities formed by clustering) | No | Yes | Yes | No | No | No | No | Yes | No | No | Yes(first 80% of the reatings for training, rest to test) | No | No | No | No | No | Yes("after which, we save the data that no less than 10 check-in records of users and POIs as a standard dataset") | No | Ranking | Precision, Recall, Average TIme | No(USPB, USG) | No | No | Yes(USG) | No | Check-ins | No | No | No | Yes | No | Yes(http://www.yelp.com/dataset_change) | Yes | Prev | No | Yelp: http://www.yelp.com/dataset_change | |||||||||||||||||||||
264 | 2018 | FCDST(for the five aspects approach and No acronym for the MF) | Exploiting the roles of aspects in personalized POI recommender systems | Baral, R., Li, T. | https://link.springer.com/article/10.1007%2Fs10618-017-0537-7 | Data Mining and Knowledge Discovery | Journal | 1,3 | 12 | \cite{DBLP:journals/datamine/BaralL18} | No | Yes(for the mf approach) | No | No | No | Yes(for the FCDST approach) | No | No | No | No | Yes(both FCDST and MF) | Yes(both FCDST and MF) | Yes(both FCDST and MF) | No | No | Yes(both FCDST and MF) | Yes | No | No | No | No | No | No | Yes( 5-Fold cross validation) | No | Yes(incomplete records removed) | No | Ranking | Precision and Recall | No(USG, LFBCA and LSBNRank) | No | Yes(MF) | Yes(USG) | No(5-fold validation) | Check-ins | No | Yes | No | No | Weeplaces | No(They refer to Personalized point-of-interest recommendation by mining users’ preference transition 2013) | Yes | Prev-No filtering | No | Weeplaces and Gowalla: Personalized point-of-interest recommendation by mining users’ preference transition 2013 | |||||||||||||||||||||
265 | 2018 | TenMF | Collaborative location recommendation by integrating multi-dimensional contextual information | Yao, L., Sheng, Q.Z., Wang, X., Zhang, W.E., Qin, Y. | https://dl.acm.org/citation.cfm?doid=3185332.3134438 | TOIT | Journal | 1 | 21 | \cite{DBLP:journals/toit/YaoSWZQ17} | Although the bibtext say 17 for 2017, the paper is well stated in 2018 | Yes | No | Yes | No | No | No | No | No | No | No | Yes | Yes | No | No | No | Yes | Yes | No | No | No | No | No | Yes(20% of the locations to test for each user randomly) | No | No | Yes(remove users and POIs with less than 10 Check-ins) | No | Ranking | Precision and Recall | No(NMF, UCF, ICF, FA, GA, LIM) | No | Yes(UB, ICF) | Yes(GMM) | No | POIs | No | Yes | No | No | Brightkite | No(They refer to Friendship and mobility: User movement in location-based social networks 2011) | Yes | Post | No | Gowalla and Brightkite: they refer to Friendship and mobility: User movement in location-based social networks 2011 | ||||||||||||||||||||
266 | 2018 | GeoEISo | A personalized point-of-interest recommendation model via fusion of geo-social information | Gao, R., Li, J., Li, X., Song, C., Zhou, Y. | https://www.sciencedirect.com/science/article/pii/S0925231217313723 | Neurocomputing | Journal | 1,2 | 33 | \cite{DBLP:journals/ijon/GaoLLSZ18} | Yes | No | Yes | Yes(Kernel estimator) | No | No | No | Gradient descent | No | No | Yes | Yes | No | No | No | No | Yes | No | No | No | No | Yes (Random 80% 20% CC repeated 5 times at dataset level) | No | No | No | Yes(users that visti at least 5 locations) | Yes(New York and Los Angeles) | Ranking | Precision and Recall | No(USG, IRenMF, MGM, CoReSo, SVD++, TrustMF, iGSLR, UA, DRW, NCP) | No | Yes(SVD) | Yes(IrenMF, others) | No | Check-ins | No | No | Yes | No | No | No(They refer to Location-based and preference-aware recommen- dation using sparse geo-social networking data 2012) | Yes | Prev-No filtering | No | Foursquare: They refer to Location-based and preference-aware recommen- dation using sparse geo-social networking data 2012 | |||||||||||||||||||||
267 | 2018 | TSG | Trust-distrust-aware point-of-interest recommendation in location-based social network | Zhu, J., Ming, Q., Liu, Y. | https://link.springer.com/chapter/10.1007/978-3-319-94268-1_58 | WASA | Conference | 1 | 2 | \cite{DBLP:conf/wasa/ZhuML18} | Yes(part of the hybrid approach) | No | Yes(They say the exploit a power law) | No | Yes(local trust propagation as part of the hybrid approach) | Yes(preferences + geo + social influence) | No | No | No | Yes | Yes | No | No | No | No | Yes | No | No | No | No | Yes(80% training, 20% test) | No | No | No | No | No | Ranking | Precison and Recall | No(CF, GD, USG) | No | Yes(U) | Yes(USG, GD) | No | Check-ins | No | Yes | Yes | No | No | No | Yes | Prev-No filtering | No | Foursquare and Gowalla: no further information | ||||||||||||||||||||||
268 | 2018 | ABPR | ABPR-- A new way of point-of-interest recommendation via geographical and category influence | Gao, J., Yang, Y. | https://link.springer.com/chapter/10.1007%2F978-981-13-2206-8_9 | ICPCSEE | Conference | 1 | 0 | \cite{DBLP:conf/icycsee/GaoY18} | No | No | Yes | No | No | No | BPR | No | No | Yes | No | Yes | No | No | No | Yes | No | No | No Information | No Information | No Information | No Information | No Information | No Information | No | No | Ranking | Precision, Recall and F1 | No(BPR, USG, IEMF) | No | Yes(BPR) | Yes(USG) | No | No information | No | Yes | Yes | No | No | No | Yes | Prev-No filtering | No | Foursquare and Gowalla: no further information | ||||||||||||||||||||||
269 | 2018 | APPR | APPR: Additive Personalized Point-of-Interest Recommendation | Naserianhanzaei, E., Wang, X., Dahal, K. | GLOBECOM | Conference | 1 | 0 | \cite{DBLP:conf/globecom/Naserianhanzaei18} | No | No | Yes | No | No | No | No | No | No | Yes | No | Yes | No | Yes(transition probabilities) | Yes(day of the week, hour of the day) | Yes | No | No | Yes(8 months for training 2 months for test) | No | No | No | No | No | No | Yes(New York and Tokyo) | Ranking | Precision, Recall and Accuracy | No(Naive Bayes and Joint) | No | No | No | No(although they say that they divide the dataset of each user in the growset and the validation set) | Check-ins | No | No | Yes | No | No | Yes(https://sites.google.com/site/yangdingqi/home/foursquare-dataset and also refer to Modeling user activity preference by leveraging user spatial temporal characteristics in lbsns 2015) | Yes | Prev-No filtering | No | Foursquare: https://sites.google.com/site/yangdingqi/home/foursquare-dataset and also refer to Modeling user activity preference by leveraging user spatial temporal characteristics in lbsns 2015. Stats does not match | |||||||||||||||||||||||
270 | 2018 | DeepRec | A deep point-of-interest recommendation system in location-based social networks | Wang, Y., Zhong, Z., Yang, A., Jing, N. | https://link.springer.com/chapter/10.1007/978-3-319-93803-5_51 | DMBD | Conference | 1 | 0 | \cite{DBLP:conf/dmbd/WangZYJ18} | No | No | No | Yes | No | No | No | Yes(DBSCAN) | No(They use embeddings but no further details) | Yes(latitudes and longitudes. Fig 2) | Yes | No | No | No | Yes(Figure 2) | Yes | No | No | No | No | Yes(70% for training, 30% for "validation") | No | No | No | Yes(remove users with less than 15 Check-ins and 5 for Brighkite) | No | Ranking | Precision and Recall | Yes(Popular, SVD, US, USG) | Yes(pOP) | Yes(US, SVD) | Yes(USG) | No | Check-ins | No | Yes | No | No | Brightkite | No(They refer to Friendship and mobility: user movement in location-based social networks 2011) | Yes | Post | No | Gowalla and Brighkite: They refer to Friendship and mobility: user movement in location-based social networks 2011 | ||||||||||||||||||||||
271 | 2018 | GeoIE | Exploiting POI-specific geographical influence for point-of-interest recommendation | Wang, H., Shen, H., Ouyang, W., Cheng, X. | https://www.ijcai.org/Proceedings/2018/539 | IJCAI | Conference | 1 | 49 | \cite{DBLP:conf/ijcai/WangSOC18} | Yes | No | Yes(part of the algorithm) | Yes(power law) | No | No | No(preferences + geographical influence, but it is not the last part, these "parts" are integrated in Eq 4) | Stochastic Gradient Ascent | No | No | Yes | No | No | No | No | No | Yes | No | No | No | Yes(70% of Check-ins as training, 15% as validation and 15% test for each user. Users Check-ins ordered chronologically) | No | No | No | No | Yes(removed users and POIs with less than 10 Check-ins) | No | Ranking | Precision and Recall | No(UCF+G, MGM+PFM, GeoMF, ankGeoFM, Geo-Teaser) | No | No | Yes(GeoMF, RankGeo) | Yes | Check-ins | No | Yes | Yes | No | No | No(They refer to Friendship and mobility: User movement in location-based social networks 2011 for Foursquare and Time-aware point-of-interest recommendation 2013 for Gowalla) | Yes | Post | No | Foursquare and Gowalla: They refer to Friendship and mobility: User movement in location-based social networks 2011 for Foursquare and Time-aware point-of-interest recommendation 2013 | |||||||||||||||||||||
272 | 2018 | ImSoRec | Exploiting implicit social relationship for point-of-interest recommendation | Zhu, H., Zhao, P., Li, Z., Xu, J., Zhao, L., Sheng, V.S. | https://link.springer.com/chapter/10.1007/978-3-319-96893-3_21 | APWeb-WAIM | Conference | 1 | 0 | \cite{DBLP:conf/apweb/ZhuZLXZS18} | No(they use similar formulations of KNN but no for Knn) | Yes(it is based on PMF, but this is more probabilistic, I think) | Yes(it is based on PMF, but this is more probabilistic, I think) | No | No | Yes(Adding geographical influence... If... Eq 15 is hybrid. I think yes) | Gradient descent | No | No | Yes | Yes | No | No | No | No | Yes | No | No | No | No | No | Yes (For each user, random 70% for training. Repeated 5 times) | No | No | Yes(removed users who have less than two concurrences. Only data from US) | Yes(United States) | Ranking | Precision and Recall | No(PMF, RegPMF, USG, SoDimRec, ASMF-LA) | No | No(PMF) | Yes(USG) | No(parameters selected according to cross validation) | Check-ins | No | Yes | Yes | No | No | No(They refer to Point-of-interest recommendations: learning potential check-ins from friends 2016 for Gowalla and Geo-SAGE: a geo- graphical sparse additive generative model for spatial item recommendation 2015) | Yes | Post | No | Foursquare and Gowalla: Geo-SAGE: a geo- graphical sparse additive generative model for spatial item recommendation 2015 for Foursquare and Point-of-interest recommendations: learning potential check-ins from friends 2011 for Gowalla | ||||||||||||||||||||||
273 | 2018 | CGA | Exploiting context graph attention for POI recommendation in location-based social networks | Zhang, S., Cheng, H. | https://link.springer.com/chapter/10.1007/978-3-319-91452-7_6 | DASFAA | Conference | 1 | 8 | \cite{DBLP:conf/dasfaa/ZhangC18} | No | No | No | Yes | Yes(Context graph as input to the model) | No | Stochastic Gradient Descent | No | Yes(They use embeddings in DNN and if only use it there, it should be no, but i do not know if it obtain the embeddings somewhere else) | Yes | Yes | No | No | No | No | Yes | No | No | No | Yes(for each user, 80% eearlist visited POIs to train, rest to test) | No | No | No | No | Yes(users with less than 15 visited pois and pois with less than 10 visited, removed) | Yes(California and Nevada) | Ranking | Precision and Recall | No(ACF, ACF+MLP, NCF, PACE) | No | No | No | No | POIs | No | Yes | Yes | No | No | No(They refer to Friendship and mobility: user movement in location-based social networks 2011 and Personalized ranking metric embedding for next new POI recommendation 2015 for Gowalla and Learning graph-based POI embedding for location-based recommendation 2017 for FS) | Yes | Post | No | Foursquare and Gowalla: They refer to Friendship and mobility: user movement in location-based social networks 2011 and Personalized ranking metric embedding for next new POI recommendation 2015 for Gowalla and Learning graph-based POI embedding for location-based recommendation 2016 for Foursquare | ||||||||||||||||||||||
274 | 2018 | Joint Model (It is not an acronym) | Personalized POI recommendation model in LBSNs | Guo, Z., Changyi, M. | https://link.springer.com/chapter/10.1007/978-3-319-69096-4_85 | IISA | Conference | ??? | 1 | 0 | \cite{Zhong2018} | No | Yes(latent factor variables) | Yes | No | No | No | No | No | No | Yes | No | No | No | No | Yes | Yes | No | No | No | No | Yes(70% for training, rest for test) | No | No | No | No | No | Ranking | Precision | No(PMF, GT, GLDA) | No | No(PMF) | No | No | Check-ins | No | No | Yes | No | No | No | None | No | Foursquare and Twitter: no further information | ||||||||||||||||||||||
275 | 2018 | ---No-Acronym-- | Point-of-interest recommendation using heterogeneous link prediction | Pourali, A., Zarrinkalam, F., Bagheri, E. | https://openproceedings.org/2018/conf/edbt/paper-320.pdf | EDBT | Conference | 1 | 3 | \cite{DBLP:conf/edbt/PouraliZB18} | No | No | No | No | Yes(link prediction) | No | No | Yes | No | Yes | Yes(friends) | Yes | No | No | No(They discuss about time intervals but i do not see more) | Yes | No | No | No | No | No | Yes(randomly select for each user a 70% of the Check-ins to the training set) | No | No | No | Yes(Austin, Chicago, Houston, Los Angeles, San Francisco) | Ranking | Precision, Recall and F1 | No(BasicMF, GeoCF, MGMMF, Markov, ML, CPOIR) | No | Yes(MF) | Yes(MGMMF) | No | Check-ins | No | Yes | No | No | No | No(They refer to Personalized point- of-interest recommendation by mining users’ preference transition 2013) | Not complete (only users stated) | None | No | Gowalla: Personalized point- of-interest recommendation by mining users’ preference transition 2013 | ||||||||||||||||||||||
276 | 2018 | ---No-Acronym-- | Exploiting spatial and temporal for point of interest recommendation | Chen, J., Zhang, W., Zhang, P., Ying, P., Niu, K., Zou, M. | https://www.hindawi.com/journals/complexity/2018/6928605/ | Complexity | Journal | 1 | 6 | \cite{DBLP:journals/complexity/ChenZZYNZ18} | Yes(part of the hybrid approach) | No | Yes(part of the hybrid approah) | No | No | Yes(Probabilistic and UBKnn with temporal influence) | No | No | No | Yes | No | Yes | No | No | Yes | Yes | No | No | No | No | No | Yes(for each user 70% or 80% of the observed data for training) | No | No | Yes(remove users and POIs with less than 5 Check-ins) | Yes(Singapore and Austin) | Ranking | Precision and Recall | No(UBCF, USG, LRT, UTE-SE, STELLAR, FPMC,TRM,TICRec) | No | Yes(UB) | Yes(STELLAR) | No | Check-ins | No | Yes | Yes | No | No | No(They refer to Friendship and mobility: user movement in location-based social networks 2011 for Gowalla and Time-aware point-of-interest recommendation 2013 for Foursquare | Yes | Post | No | Foursquare and Gowalla: Time-aware point-of-interest recommendation 2013 | ||||||||||||||||||||||
277 | 2018 | ST-DME | Spatial-temporal distance metric embedding for time-specific POI recommendation | Ding, R., Chen, Z., Li, X. | https://ieeexplore.ieee.org/document/8528314 | IEEE Access | Journal | 1 | 12 | \cite{DBLP:journals/access/DingCL18} | Time Aware | No | No(metric embedding is deep learning) | No | Yes | No | Yes(Temporal, geographical) | stochastic gradient decent (SGD) | Yes | Yes(metric embedding) | Yes | No | No | No | Yes | Yes | Yes | No | No | No | Yes(ordering the Check-ins by timestamp and for each user, 70% fon training, 10% for validation and 20% to test) | No | No | No | No | Yes(remove users that have visited less than 5 POIs and POIs visited by less than 5 users) | Yes(Tokyo, New York) | Ranking | Accuracy(HitRatio) and MRR | Yes*(Time-Pop, BPR-MF, UTS) | Yes(Pop) | Yes(BPR) | Yes(RankGeo) | Yes(10% for each user for validation) | Check-ins | No | Yes | Yes | No | No | No(They refer to A sentiment-enhanced person- alized location recommendation system 2013 and Modeling user activity preference by leveraging user spatial temporal characteristics in LBSNs 2015 for Foursquare and Personalized point-of-interest recommendation by mining users’ preference transition 2013 Exploiting geographical neigh- borhood characteristics for location recommendation 2014 for Gowalla) | Yes | Post | No | Foursquare and Gowalla: they refer to A sentiment-enhanced person- alized location recommendation system 2013 and Modeling user activity preference by leveraging user spatial temporal characteristics in LBSNs 2015 and Personalized point-of-interest recommendation by mining users’ preference transition 2013 and ‘Exploiting geographical neigh- borhood characteristics for location recommendation 2014 for Gowalla | |||||||||||||||||||||
278 | 2018 | TSG-list MF | A list-wise matrix factorization based POI recommendation by fusing multi-tag, social and geographical influences | Zhang, Z., Liu, Y. | https://jit.ndhu.edu.tw/article/view/1632 | Journal of Internet Technology | Journal | ??? | 1 | 2 | \cite{Zhang2018} | No(For modeling the social influence they state something similar to UB, similarities are not collaborative) | Yes | No | No | No | Yes(it is a fusion approach) | Gradient based approaches | No | No | Yes | Yes | Yes(tags) | No | No | No | Yes | No | No | No | No | No | Yes | No | Yes | Yes(remove users with less than 20 ratings) | No | Ranking | HitRatio, MRR, Recall | No(MF, NMF, List MF, T-list MF, S-list MF, G-list MF) | No | Yes(MF) | Yes(G-list) | No(Cross validation) | POIs | No | No | No | Yes | No | No | Yes | Post | No | Yelp: no further information | |||||||||||||||||||||
279 | 2018 | LBIMC | Noise-tolerance matrix completion for location recommendation | Xia, B., Li, T., Li, Q., Zhang, H. | https://link.springer.com/article/10.1007%2Fs10618-017-0516-z | Data Mining and Knowledge Discovery | Journal | 1 | 8 | \cite{DBLP:journals/datamine/XiaLLZ18} | No | Yes(It is matrix completion but I think it fits as MF) | No | No | No | No | stochastic proximal gradient descent (SPGD) | No | No | No | No | No | No | No | No | Yes | No | No | Yes(3 months for training 1 for test. 10 partitions) | No | No | No | No | No | No | Yes(Manhattan) | Ranking | Precision, Recall and Coverage | No(UserCF, ItemCF, MP, NMF, PMF, BPR, ESSVM) | No | Yes(UCF, ICF, BPR) | Yes(ESSVM) | No | Check-ins - Tweets | No(they discuss about it) | No | Yes | No | No | No | Yes | Prev-No filtering | No | Foursquare using Twitter: no further info provided | ||||||||||||||||||||||
280 | 2018 | MFRA | A Multi-factor Recommendation Algorithm for POI Recommendation | Yang, R., Han, X., Zhang, X. | https://link.springer.com/chapter/10.1007%2F978-3-030-02934-0_41 | WISA | Conference | 1 | 2 | \cite{DBLP:conf/IEEEwisa/YangHZ18} | Yes(part of the hybrid approach) | No | No | No | No | Yes(Social and preferences, also consider the popularity of the venues) | No | No | No | No | Yes | Yes | No | No | No | Yes | No | No | No | No | Yes(75% of the dataset as training and rest to test. I would vote for random CC) | No | No | No | No | Yes(New York and Los Angeles) | Ranking | Precision and Recall | No(UCF, FCF, USG) | No | Yes(UCF) | Yes(USG) | No | Check-ins | No | No | Yes | No | No | No | Yes | Prev-No filtering | No | Foursquare: no further info provided | ||||||||||||||||||||||
281 | 2018 | TMCA | Next Point-of-Interest Recommendation with Temporal and Multi-level Context Attention | Li, R., Shen, Y., Zhu, Y. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8594953 | ICDM | Conference | 1 | 19 | \cite{DBLP:conf/icdm/LiSZ18} | Next POI recommendation | No | No | No | Yes | No | No | No | No | No(use embeddings but in the DNN) | Yes | No | Yes | No | Yes | Yes | Yes | No | No | No | Yes(70%-train 20%valdiation 20% test for each user) However, it sums 110%??? | No | No | No | No | Yes(remove users and items with less that 10 ratings) | No | Ranking | Recall and NDCG | Yes(Pop, FPMC, PRME, RNN, LSTM and variants of the proposed algorithm) | Yes(Pop) | No | Yes(PRME) | Yes | Check-ins | No | Yes | No | Yes | No | Yes(http://snap.stanford.edu/data/loc-gowalla.htm and https://www.yelp.com/dataset/challenge) | Yes | Prev | Yes(https://github.com/zhenql/TMCA) | Yes | No | Yes | Gowalla and Yelp: http://snap.stanford.edu/data/loc-gowalla.htm for Gowalla and https://www.yelp.com/dataset/challenge for Yelp. They also refer to Bridging Collaborative Filtering and Semi-Supervised Learning: A Neural Approach for POI Recommendation 2017 | ||||||||||||||||||
282 | 2018 | JLGE | Recommendation of points-of-interest using graph embeddings | Christoforidis, G., Kefalas, P., Papadopoulos, A., Manolopoulos, Y. | https://ieeexplore.ieee.org/document/8631456 | DSAA | Conference | 1 | 8 | \cite{DBLP:conf/dsaa/ChristoforidisK18} | Time Aware | No | Yes(graph embedding) | No(negative sampling for optimization) | No | No | No(Temporal, Learning factors using temporal and social factors. I think yes) | No | No | Yes(graph embedding) | Yes | Yes | No | No | No | Yes | Yes | No | No | No | No | Yes(80% training, 10% validation, 10% test) | No | No | No | No | No | Ranking | Accuracy | No(GE) | No | No | Yes(GE) | Yes(10%) | Check-ins | No | No | Yes | No | Weeplaces | Yes(They refer to Learning graph-based poi embedding for location-based recommendation 2016 for Foursquare and “MAPS: A multi aspect personalized poi re- commender system 2016 for Weeplaces) | Yes | Prev-No filtering | Yes(https://github.com/thedx4/JLGE) | Yes | No | No | Foursquare and Weeplaces: https://sites.google.com/site/dbhongzhi 2http://www.yongliu.org/datasets for Foursquare and http://www.yongliu.org/datasets for Weeplaces.They also refer to “Learning graph-based poi embedding for location-based recommendation 2016 for foursquare and MAPS: A multi aspect personalized poi re- commender system for Weeplaces | ||||||||||||||||||
283 | 2018 | TCENR | TCENR: A hybrid neural recommender for location based social networks | Tal, O., Liu, Y. | https://ieeexplore.ieee.org/document/8637404 | ICDM | Conference | 1 | 1 | \cite{DBLP:conf/icdm/TalL18} | Move this paper to 2018 | No | No | No | Yes | No | No | Gradient Descent | No | Yes(word embedding) | No | No | No | Yes(text reviews) | No | No | Yes | No | No | No | No | Yes(56% training, 24% validation and 20% test) | No | No | No | Yes(all users and locations with less than 100 written reviews removed and less than 10 friends) | No | Error/Ranking | MSE, Precision and Recall | No(HPF, NeuMF, PACE, DeepCoNN) | No | No | No | Yes(24% validation) | POIs(Equivalent) | No | No | No | Yes | No | Yes(https://www.yelp.com/dataset/challenge) | No | None | No | Yelp: https://www.yelp.com/dataset/challenge | |||||||||||||||||||||
284 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
285 | 2019 | ??? | Effective knowledge based recommender system for tailored multiple point of interest recommendation | Vijayakumar, V., Vairavasundaram, S., Logesh, R., Sivapathi, A. | IJWP | Journal | 1 | 19 | IGNORE | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
286 | 2019 | ??? | Swarm intelligence clustering ensemble based point of interest recommendation for social cyber-physical systems | Devarajan, M., Fatima, N.S., Vairavasundaram, S., Ravi, L. | Journal of Intelligent and Fuzzy System | Journal | 1 | 12 | IGNORE | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
287 | 2019 | ??? | HybRecSys: Content-based contextual hybrid venue recommender system | Bozanta, A., Kutlu, B. | https://journals.sagepub.com/doi/10.1177/0165551518786678 | Journal of Information Science | Journal | 1 | 2 | IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
288 | 2019 | ??? | Points of interest recommendations based on check-in motivations | Vakeel, K.A., Ray, S. | Tourism Analysis | Journal | ??? | 1 | 0 | IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
289 | 2019 | ??? | Fused collaborative filtering with user preference, geographical and social influence for point of interest recommendation | Zeng, J., Li, F., He, X., Wen, J. | https://www.igi-global.com/gateway/article/238000 | nternational Journal of Web Services Research | Conference | 1 | 3 | IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
290 | 2019 | ??? | Point-of-Interest Recommendation Based on User Contextual Behavior Semantics | Yu, D., Xu, K., Wang, D., Yu, T., Li, W. | https://www.worldscientific.com/doi/abs/10.1142/S0218194019400217 | International Journal of Software Engineering and Knowledge Engineering | Journal | 1 | 2 | ??? | IGNORE. Pdf not available | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
291 | 2019 | ??? | Point-of-Interest Recommendation in Location-Based Social Networks | Cakmak, E., Kaya, B., Kaya, M. | https://ieeexplore.ieee.org/document/8965501 | International Informatics and Software Engineering Conference | Conference | ??? | 1 | 0 | ??? | IGNORE. Very poorly written | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
292 | 2019 | LIOR | A location and intention oriented recommendation method for accuracy enhancement over big data | Rafique, W., Qi, L., Zhou, Z., Zhao, X., Tang, W., Dou, W. | https://link.springer.com/chapter/10.1007%2F978-3-030-28468-8_1 | MobiCASE | Conference | 1 | 0 | They do not recommend POIs, they use the location but the dataset is Movielens. IGNORE --> different task: they use location for recommendation | Yes(part of the hybrid approach - IBKNN | No | No | No | No | Yes | Yes | No | No | No | No | No | No | Yes(80% of the data for training, rest to test) | No | No | No | No | Accuracy, MAP | No(ULA-LDA, MLTRS, LARS) | No | No information (in movielens there are no repetitions) | No(They discuss about it) | No | No | No | Movielens | Yes(http://grouplens.org/datasets/movielens/) | No | |||||||||||||||||||||||||||||||||||||
293 | 2019 | PP-TRR | Toward pattern and preference-aware travel route recommendation over location-based social networks | Zhu, L., Yu, L., Cai, Z., Zhang, J. | Journal of Information Science and Engineering | Journal | 1 | 1 | Route recommendation. IGNORE | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
294 | 2019 | ---No-Acronym-- | Distributed representations of users and locations for friendship recommendation on location-based social network | Chen, Z., Zhan, Y. | IJWP | Journal | 1 | 0 | Friendship recommendation. IGNORE <-- no experiments <-- URL was not correct | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
295 | 2019 | ---No-Acronym-- | A location recommendation based on user reviews using cart | Janani, V., Balasubramanian, L., Sasikala, G., Vidhya, G., Kowsalya, T. | https://ieeexplore.ieee.org/document/8878812 | ICSCAN | Conference | ??? | 1 | 0 | It seems they perform hotels recommendation and the images and formulas in the paper are awful. IGNORE <-- recommendation task is not obvious | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
296 | 2019 | Context-aware group-oriented location recommendation in location-based social networks | Khazaei, E., Alimohammadi, A. | https://www.mdpi.com/2220-9964/8/9/406 | International Journal of Geo-Information | Journal | 1 | 3 | Group recommenation. Maybe it is not on the scope. IGNORE | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
297 | 2019 | ??? | Exploiting the user activity-level to improve the models' accuracy in point-of-interest recommender systems | Luiz Chaves, Nicollas Silva, Rodrigo Carvalho, Adriano C. M. Pereira, Leonardo C. da Rocha | https://dl.acm.org/doi/abs/10.1145/3323503.3349551 | Webmedia | Conference | 3 | 0 | It is in portuguese. IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
298 | 2019 | ---No-Acronym--- | A location-based social network system integrating mobile augmented reality and user generated content | Yuanwen Yue, Jiaqi Ding, Yuhao Kang, Yueyao Wang, Kunlin Wu,Teng Fei | https://dl.acm.org/doi/abs/10.1145/3356994.3365507 | SIGSPATIAL | Conference | 3 | 0 | No offline experiment, nor baselines. IGNORE <-- it is like a demo | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
299 | 2019 | ??? | Geographic-categorical diversification in POI recommendations | Luiz Chaves, Nicollas Silva, Rodrigo Carvalho, Adriano C. M. Pereira, Leonardo C. da Rocha | https://dl.acm.org/doi/abs/10.1145/3323503.3349554 | Webmedia | Conference | 3 | 0 | It is in portuguese. IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
300 | 2019 | DPTCR | Differential privacy-based trajectory community recommendation in social network | Jianhao Wei, Yaping Lin, Xin Yao, Voundi Koe Arthur Sandor | https://www.sciencedirect.com/science/article/pii/S0743731518309572 | Journal of Parallel Distribution Computing | Journal | 2 | 0 | Trajectory recommendation. IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
301 | 2019 | ---No-Acronym--- | A statistical approach to participant selection in location-based social networks for offline event marketing | Yuxin Liu, Anfeng Liu, Xiao Liu, Xiaodi Huang | https://www.sciencedirect.com/science/article/pii/S0020025518309654 | Information Sciences | Journal | 2 | 0 | I think it is not on the scope. Participants and events. IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
302 | 2019 | SS-ILM | Discovering socially important locations of social media users | Ahmet Sakir Dokuz, Mete Celik | https://www.sciencedirect.com/science/article/pii/S0957417417303949 | Expert Systems With Apllications | Journal | 2 | 0 | Not using LBSN. Only twitter. IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
303 | 2019 | ---No-Acronym-- | Research on comprehensive point of interest (POI) recommendation based on spark | He, F., Wei, P. | https://link.springer.com/article/10.1007/s10586-018-2061-y | Cluster Computing | Journal | ??? | 1 | 4 | Not sure if it is on the scope. They do not propose any specific algorithm. IGNORE --> it is a survey | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
304 | 2019 | ??? | Predicting attributes and friends of mobile users from AP-Trajectories | Pinghui Wang, Feiyang Sun, Di Wang, Jing Tao, Xiaohong Guan, Albert Bifet | https://www.sciencedirect.com/science/article/pii/S0020025518304699 | Information Sciences | Journal | 2 | 0 | Not POI recommendation. IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
305 | 2019 | ---No-Acronym-- | Location Based Place Recommendation using Social Network | Naik, P., Desai, P.V., Pati, S. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=9033625 | International Conference for Convergence of Technology | Conference | ??? | 1 | 1 | ??? | Survey on POIs, IGNORE | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
306 | 2019 | CTIR | Complementing travel itinerary recommendation using location-based social networks | Zhou, J., Gu, Y., Lin, W. | https://ieeexplore.ieee.org/document/9060090 | Symposia and Workshops on Ubiquitous, Autonomic and Trusted Computing | Conference | 1 | 0 | \cite{DBLP:conf/uic/ZhouGL19} | Travel recommendation. IGNORE | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
307 | 2019 | ---No-Acronym-- | Location based point-of-interest recommendation system using co-pear similarity measure | Vinodha, R., Parvathi, R. | http://www.ijstr.org/paper-references.php?ref=IJSTR-1219-27355 | International Journal of Scientific and Technology Research | Journal | ??? | 1 | 0 | The images have very bad resolution and it is very bad written. <-- although presentation is bad, the task fits our requirements, so it should be included. IGNORE | Yes | No | No | No | No | No | No | Yes(Density-based spatial clustering and Noise (DBSCAN)) | No | Yes | No | No | No | Yes | Yes | No Information | No Information | No Information | No Information | No Information | No Information | No | No | None | No | No(HITS) | No | No | No | No | No information | No(they discuss about it) | No | No | No | GeoLife | Yes(https://www.microsoft.com/en-us/research/publication/geolife-gps-trajectory-dataset-user-guide/), although it is bad reference | Not complete (only trajectories) | None | No | Geolife: reference of the paper is incorrect | ||||||||||||||||||||||||
308 | 2019 | DRPS | Dynamic recommendation of POI sequence responding to historical trajectory | Huang, J., Liu, Y., Chen, Y., Jia, C. | https://www.mdpi.com/2220-9964/8/10/433 | International Journal of Geo-Information | Journal | 1 | 4 | \cite{DBLP:journals/ijgi/HuangLCJ19} | Sequence of POIs generation. Specific POI Sequence generation and evaluation. IGNORE | No | No | No | Yes | No | No | No | No | No(Yes but in the DNN) | Yes | No | Yes | No | Yes | No | No | No | Yes(70% training, 10% validation and 20% test. They repeat the same experiment 10 times) | No | No | No | No | Yes(NewYork, San Francisco, Brooklyn and London) | Ranking | Ap and OSP(sequential precision) | No(RAND, AMC, LORE, LSTM-Seq2Seq) | Yes(Random) | No | Yes(LORE) | Yes(10%) | Check-ins | Yes(users with more than 15 Check-ins and less than 35 are cold start) | No | No | No | Weeplace | No(They refer to Personalized Point-of-interest Recommendation by Mining Users’ Preference Transition 2013) | Yes | Prev-No filtering | No | Weeplace (I suppose it is Weeplaces) and Gowalla: they refer to Personalized point-of-interest recommendation by mining users’ preference transition 2013 for Weeplaces | ||||||||||||||||||||||||
309 | 2019 | TSG | Friend and POI recommendation based on social trust cluster in location-based social networks | Zhu, J., Wang, C., Guo, X., Ming, Q., Li, J., Liu, Y. | Journal Wireless Communication and Networking | Journal | 1 | 5 | \cite{DBLP:journals/ejwcn/ZhuWGMLL19} | Friend and POI recommendation. Friend recommendation is FRTC and POI recommendation is TSG. IGNORE. Repeated in 2018 \cite{DBLP:conf/wasa/ZhuML18} | Yes | No | No | No | No | Yes(Social, geographical and collaborative) | No | Yes | No | Yes | Yes | No | No | No | No | No | No | Yes(80% of the data for training 20% for test) | No | No | No | Yes(users with less than 10 Check-ins removed and less than 5 friends and POIs with less than 5 Check-ins also removed) | Yes (New York and Austin) | Ranking | Precision and Recall | No(U, GD) | No | Yes(U) | Yes(GD) | No | Check-ins | No | Yes | Yes | No | No | No | Yes | Post | No | Foursquare and Gowalla: No further details | |||||||||||||||||||||||||
310 | 2019 | TSG-MF | Fused matrix factorization with multi-tag, social and geographical influences for POI recommendation | Zhang, Z., Liu, Y., Zhang, Z., Shen, B. | https://link.springer.com/article/10.1007%2Fs11280-018-0579-9 | World Wide Web | Journal | 1,3 | 18 | \cite{DBLP:journals/www/ZhangLZS19} | IGNORE. Repeated in \cite{Zhang2018} | No(Uses a similarity matrix) | Yes | No | No | No | No | Gradient Based approaches | Yes | No | Yes | Yes | No | No | No | No | No | No | Yes(different test percentages, 90%, 80% ,70%, 60%) | No | No | No | Yes(they preprocess the subsets, but not specific how) | No | Error/Ranking | RMSE, MAE, Precision, Recall | No(MF, T-MF, S-MF, G-MF) | No | Yes(MFs) | Yes(G-MF) | No | POIs | No | No | No | Yes | No | No | Yes | Post | No | Yelp: they process the data but no further info abot the data obtention | ||||||||||||||||||||||||
311 | 2019 | TCENR | A Joint Deep Recommendation Framework for Location-Based Social Networks | Tal, O., Liu, Y. | https://www.hindawi.com/journals/complexity/2019/2926749/ | Complexity | Journal | 1,3 | 3 | \cite{DBLP:journals/complexity/TalL19} | It is repeated in 2018. IGNORE. The same as \cite{DBLP:conf/icdm/TalL18} | No | No | No | Yes | No | No | Gradient Descent | No | Yes(word embedding) | Yes | Yes | Yes(textual information) | No | No(I think not as they only model sequential texts) | No | No | No | Yes(56%training, 24% validation and 20% test) | No | No | No | Yes(locations and users with less than 1000 reviews and less than 10 friends are removed) | No | Ranking | Accuracy and training time | No(HPF, NMF, Geo-SAGE, LCARS, NeuMF, PACE, DeepCoNN) | No | Yes(NMF) | Yes(LCARS, GeoSage) | Yes(24% validation) | POIs(Equivalent) | No | No | No | Yes | No | Yes(https://www.yelp.com/data- set/challenge) | No | None | No | Yelp: https://www.yelp.com/data- set/challenge | ||||||||||||||||||||||||
312 | 2019 | DRTL | On cross-domain transfer in venue recommendation | Manotumruksa, J., Rafailidis, D., Macdonald, C., Ounis, I. | https://link.springer.com/chapter/10.1007%2F978-3-030-15712-8_29 | ECIR | Conference | 1 | 0 | \cite{DBLP:conf/ecir/ManotumruksaRMO19} | They do not propose any specific POI recommendation approach. --> task is POI recommendation, so it should be considered <-- to be included in our cross-domain paper | No | Yes | No | Yes | No | No | BPR | No | No | No | No(For the Crossfire they use it but the proposed approach is not crossfire) | No | No | Yes | No | Yes | No | No | No | Yes(leave one out methodology) | No | No | No | No | Yes(remove venues with less than 10 interactions) | No | Ranking | HitRatio, NDCG | No(MF, DRCF) | No | Yes(MF) | No | No | Check-ins | No | No | Yes | Yes | Brightkite | Yes(https://snap.stanford.edu/data/ for Brightkite, https://archive.org/details/201309 for Foursquare and https://www.yelp.com/dataset challenge for Yelp) | Yes | Post | Yes(but URL does not work) | No | No | Foursquare, Brightkite and Yelp: https://snap.stanford.edu/data/ for Brightkite, https://archive.org/details/201309 foursquare dataset umn for FOursquare and https://www.yelp.com/dataset challenge for Yelp | |||||||||||||||||||
313 | 2019 | ---No-Acronym-- | Influence-Time-Proximity Driven Locations Recommendation Model: An Integrated Approach | Gupta, A., Tandon, N., Khetarpaul, S. | https://ieeexplore.ieee.org/document/8929590 | TENCON | Conference | 1 | 0 | \cite{DBLP:conf/tencon/GuptaTK19} | Yes(part of the hybrid apporach for friends) | No | No | No | No | Yes(seasonality and social influence) | No | No | No | No(In the image they say they exploit the distance but not clear) | Yes | Yes | No | No | Yes | Yes | No | No | No Information | No Information | No Information | No Information | No Information | No Information | No | No | Ranking | Precision, Recall, F1 | No(Only parts of the hybrid approach as baselines) | No | No | No | No | No information | No | No | Yes | No | No | No(They refer to Recommendation System for Location- based Social Network -Unpublished) | Not complete (POIs not stated) | None | No | Foursquare: Recommendation System for Location- based Social Network unpublished. Not complete statistics | ||||||||||||||||||||||
314 | 2019 | GeoSoCa and MF | Comparison of sentiment analysis and user ratings in venue recommendation | Wang, X., Ounis, I., Macdonald, C. | https://link.springer.com/chapter/10.1007%2F978-3-030-15712-8_14 | ECIR | Conference | 1 | 4 | \cite{DBLP:conf/ecir/WangOM19} | Geosoca and MF with sentiment analysys. Not sure if it is on the scope. --> task is venue recommendation, it should be considered | No | Yes(the MF). But I think, we should focus on the sentiment analysis approaches | No | Yes(part of the sentiment analysis can be deep learning) | No | Yes(for GeoSoCa Social, Categorical and Geographical). But I think, we should focus on the sentiment analysis approaches | No | No | No | Yes(GeoSoca) | Yes(GeoSoca) | Yes(categorical information) | Yes(textual information form the reviews. I only put here the information used by the sentiment analysis) | No | No | Yes | No | No | No | No | No | No | Yes (5-fold cross validation at dataset level) | No | Yes(removed users with less than 20 reviews and venues with less than 5 visits) | Yes (Phoenix and Las Vegas and a Cross-City) | Ranking | Precision, MAP | No(uses GeoSoca, MF and parts of geosoca using different sentiment analysis approaches, Rand, Rating, SWN, SVM, CNN, LSTM) | Yes(Random) | No | No | No(5-fold cross validation) | POIs(Equivalent) | No | No | No | Yes | No | Yes(https://www.yelp.co.uk/dataset/challenge) | Yes | Post | No | Yelp: https://www.yelp.co.uk/dataset/challenge | |||||||||||||||||||||
315 | 2019 | CCS-POI-RS | Context-Category Specific sequence aware Point-Of-Interest Recommender System with Multi-Gated Recurrent Unit | Kala, K.U., Nandhini, M. | https://link.springer.com/article/10.1007/s12652-019-01583-w | Journal of Ambient Intelligence and Humanized Computing | Journal | ??? | 1 | 2 | \cite{Kala2019} | No | No | No | Yes | No | No | Stochastic Gradient Descent | No | Yes(but in the DNN) | Yes | No | No | No | Yes | Yes | Yes | No | No | No | Yes(leave one out methodology) | No | No | No | No | Yes(removed POIs with less than 10 Check-ins) | Yes (Austin for Gowalla and Singapore for Foursquare) | Ranking | Recall and NDCG | No(STGN, CAGRU, TimeGRU, CGRU, LatentCross) | No | No | No | No | Check-ins but the selected 100 are venues, not Check-ins | Yes | Yes | Yes | No | No | No(They refer to Friendship and mobility: user movement in location-based social networks 2011 for Gowalla and Time-aware point-of-interest recommendation 2013 for Foursquare) | Yes | No information | No | Foursquare and Gowalla: They refer to Friendship and mobility: user movement in location-based social networks 2011 for Gowalla and Time-aware point-of-interest recommendation 2013 for Foursquare. Wrong reported statistics. The statistics reported from Foursquare are from Gowalla | |||||||||||||||||||||
316 | 2019 | HWREC | Context-Aware Point-of-Interest Recommendation Algorithm with Interpretability | Zhang, G., Qi, L., Zhang, X., Xu, X., Dou, W. | https://link.springer.com/chapter/10.1007%2F978-3-030-30146-0_50 | CollaborateCom | Conference | 1 | 0 | \cite{DBLP:conf/colcom/ZhangQZXD19} | Yes(although it is not a classical knn. THey use a CF to select candidates POIs) | No | No | No | No | No | Yes(Hawkess process) | No | Yes | No | Yes | No | No | No | Yes | Yes | Yes | No | No | No | Yes(for each user 80% of ordered Check-ins to training) | No | No | No | No | Yes(For Foursquare removed users and POIs with less than 10 Check-ins. For Gowalla, removed users with less than 5 Check-ins and POIs visited less than 10 times) | Yes (New York for Gowalla and Tokyo for Foursquare) | Ranking | Precision and Recall | No(HUFF, AMC, ASVD++, AOBPR, LORE) | No | No | Yes(LORE) | No | Check-ins | No | Yes | Yes | No | No | Yes for Gowalla http://snap.stanford.edu/data/loc-gowalla.html and For Foursquare, they refer to Modeling user activity preference by leveraging user spatial temporal characteristics in LBSNs 2015 | Yes | Post | No | Gowalla and Foursquare: http://snap.stanford.edu/data/loc-gowalla.html for Gowalla and Modeling user activity preference by leveraging user spatial temporal characteristics 2015 for Foursquare | |||||||||||||||||||||
317 | 2019 | L-WMF | Location perspective-based neighborhood-aware POI recommendation in location-based social networks | Guo, L., Wen, Y., Liu, F. | https://link.springer.com/article/10.1007%2Fs00500-018-03748-9 | Soft Computing | Journal | 1 | 13 | \cite{DBLP:journals/soco/0008WL19} | Yes(It seems they compute similarities between the venues) | Yes | Yes(Probabilistic MF) | No | No | No | Alternate Least squares | No | No | Yes | No | No | No | No | No | Yes | No | No | No | Yes( for each user 70% training 10% validation and 20% test) | No | No | No | No | Yes(removed users with less than 15 Check-ins and POIs with less than 10 Check-ins) | Yes(USA for Foursquare) | Ranking | Precision and Recall | Yes(Pop, LRT,MGMPFM,iGSRL,BPRMF,WMF) | Yes(Pop) | Yes(WMF) | Yes(IGSLR, MGMPFM) | Yes(10%) | Check-ins | No | Yes | Yes | No | No | Yes(http://spatialkeyword.sce.ntu.edu.sg/eval-vldb17/) | Yes | Post | No | Foursquare and Gowalla: http://spatialkeyword.sce.ntu.edu.sg/eval-vldb17/ for both | ||||||||||||||||||||||
318 | 2019 | ---No-Acronym-- | A novel next new point-of-interest recommendation system based on simulated user travel decision-making process | Jiao, X., Xiao, Y., Zheng, W., Wang, H., Hsu, C.-H. | Future Generation Computer Systems | Journal | 1,2 | 11 | \cite{DBLP:journals/fgcs/JiaoXZWH19} | No | Yes | No(part of the hybrid approach) | No | No | Yes(travel,¡ + preference probability) | No | Yes | No | Yes | No | No | No | Yes | Yes | Yes | No | No | Yes(last two months to test, rest to train) | No | No | No | No | No | Yes(remove consecutive Check-ins with unrealistic speed and only users with at least 4 Check-ins per week) | Yes(New York and Tokyo) | Ranking | Precision, Recall and F1 | No(UCF, PME, LORE, FPMC-LR, PRME) | No | Yes(UCF) | Yes(LORE) | No | Check-ins | No | Yes | Yes | No | No | Yes(https://sites.google.com/site/yangdingqi/home/foursquare-dataset for Foursquare and http://snap.stanford.edu/data/loc-Gowalla.html for Gowalla) | Yes | Post | No | Foursquare and Gowalla: https://sites.google.com/site/yangdingqi/home/foursquare-dataset for Foursquare (They also refer to Modeling user activity preference by leveraging user spatial temporal characteristics in LBSNs 2014) and http://snap.stanford.edu/data/loc-Gowalla.html for Gowalla (They also refer to Friendship and mobility:user movement in location-based social networks 2011) | |||||||||||||||||||||||
319 | 2019 | UFC | UFC: A Unified POI Recommendation Framework | Zhou, J., Liu, B., Chen, Y., Lin, F. | https://link.springer.com/article/10.1007/s13369-019-04011-5 | Arabian Journal for Science and Engineering | Journal | ??? | 1 | 2 | \cite{Zhou2019} | Yes(part of the hybrid approach) | No | Yes(part of the hybrid approach, by distance) | No | Yes(part of the hybrid approach) | Yes(Social, geographical and CF combined) | No | No | No | Yes | Yes | No | No | No | No | Yes | No | No | No Information | No Information | No Information | No Information | No Information | No Information | No | No | Ranking | Precision, Recall, NDCG and Map | No(MGM,LFBCA,LORE,FSBPR,SRMP) | No | No | Yes(IGSLR, LORE) | No | No information | No | Yes | No | Yes | No | No | Yes | Prev-No filtering | No | Gowalla and Yelp: no further details | |||||||||||||||||||||
320 | 2019 | AKAWO | Recommendations based on user effective point-of-interest path | Zhou, G., Zhang, S., Fan, Y., Li, J., Yao, W., Liu, H. | https://link.springer.com/article/10.1007%2Fs13042-018-00910-5 | International Journal Machine Learning & Cybernetics | Journal | 1 | 3 | \cite{DBLP:journals/mlc/ZhouZFLYL19} | No | No | No | No | No | Yes(Eq 4) | Yes(Approximate Knapsack Algorithm with Optimization) | No | No | No | No | No | Yes(They discuss about features) | No | No | Yes | Yes | No | No | No | No | No | Yes(80% of visited POIs to training, rest to test | No | No | Yes(remove users and POIs with less than 5 Check-ins) | No | Ranking | Precision, Recall and DIversity | No(UST, GA) | No | No | Yes(UTESE) | No | POIs | No | Yes | Yes | No | No | No | Yes | Post | No | Foursquare and Gowalla: no further details | |||||||||||||||||||||
321 | 2019 | HiRecS | HiRecS: A Hierarchical Contextual Location Recommendation System | Baral, R., Iyengar, S.S., Zhu, X., Li, T., Sniatala, P. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8848605 | IEEE Transactions on Computational Social Systems | Journal | 1 | 4 | \cite{DBLP:journals/tcss/BaralIZLS19} | Sequence of POIs generation. But it also has normal POI recommendation | No | No | Yes | No | No(It is not a social graph, but the hierarchy could not be considered as a graph) | Yes(Hierarchical aggregation (also use clustering)) | Yes(Hierarchical aggregation (also use clustering)) | Expectation maximization algorithm | Yes | No | Yes | Yes | Yes | No | No | Yes | Yes | No | No | No | No | No | No | Yes (5-fold cross validation) | No | No | No | Ranking | Precision, Recall, NDCG, Displacement, Diversity, F1 | Yes(Pop, UCF, UCF+G, GeoMF,HSR,ASMF-LA,HGMF) | Yes(Pop) | Yes(UCF) | Yes(UF-G, GeoMF) | No(5-fold cross validation) | Check-ins | No | Yes | No | No | Weeplace | Yes(http://www.yongliu.org/datasets/ and they refer to Personalized point-of-interest recommendation by mining users’ preference transition 2013) | Yes | Prev-No filtering | No | Weeplace (I suposse it is Weeplaces) and Gowalla: they refer to Personalized point-of-interest recommendation by mining users’ preference transition 2013 and http://www.yongliu.org/datasets/ | ||||||||||||||||||||
322 | 2019 | GPDM and PPDM | Next and next new POI recommendation via latent behavior pattern inference | Li, X., Han, D., He, J., Liao, L., Wang, M. | https://dl.acm.org/citation.cfm?doid=3357218.3354187 | TOIS | Journal | 1,3 | 9 | \cite{DBLP:journals/tois/LiHHLW19} | Next POI recommendation | No | Yes(I think yes they say the use latent features. Expectation maximization algorithm, TUcker descomposition) | Yes(for both methods, power law + BPR) | No | No | No | Expectation maximization algorithm, BPR | No | No | Yes | No | Yes(for both) | No | Yes(for both) | No | Yes | No | No | No | Yes(80% training for each user and 20% test, ordered) | No | No | No | No | Yes(users with less than 10 Check-ins removed) | Yes (for Foursquare, New York and Los Angeles, for Gowalla, not) | Ranking | Precision, Recall and NDCG | No(MF, PMF,FPMC-LR,PRME-G, GeoSoca,Rank-GeoFm, ST-RNN) | No | Yes(MF, PMF) | Yes(Rank-GeoFm) | No | Check-ins | No | Yes | Yes | No | No | No(They refer to Location-based and preference-aware recommendation using sparse geo-social networking data 2012 for FS and Fused matrix factorization with geographical and social influence in location-based social networks 2012 for Gowalla) | Yes | Post | No | Foursquare and Gowalla: Location-based and preference-aware recommendation using sparse geo-social networking data 2012 for Foursquare and Fused matrix factorization with geographical and social influence in location-based social networks 2012 for Gowalla | |||||||||||||||||||||
323 | 2019 | CATAPE | Category-aware location embedding for point-of-interest recommendation | Rahmani, H.A., Aliannejadi, M., Zadeh, R.M., Baratchi, M., Afsharchi, M., Crestani, F. | https://dl.acm.org/citation.cfm?doid=3341981.3344240 | ICTIR | Conference | 1,3 | 4 | \cite{DBLP:conf/ictir/RahmaniAZBAC19} | No | No | No | Yes(they say it is graph embeggin but define as neural model) | No | No | No | No | Yes | No | No | Yes | No | Yes | No | Yes | No | No | No | Yes(80% of the Check-ins for training, rest to test) | No | No | No | No | No | No | Ranking | Precision and Recall | No(USG, MGMPFM,BPRMF,RankGEoFM,HGMF, Metric factorization) | No | Yes(BPRMF) | Yes(MGMPFM, Rank-GeoFm) | No | Check-ins | No | Yes | No | Yes | No | No(They refer to An experimental evaluation of point-of-interest recommendation in location-based social networks 2017 for Yelp and Exploiting geo- graphical influence for collaborative point-of-interest recommendation 2011 for Gowalla) | Yes | Prev-No filtering | No | Gowalla and Yelp: no further details | ||||||||||||||||||||||
324 | 2019 | MEAP-T | Time-aware metric embedding with asymmetric projection for successive POI recommendation | Ying, H., Wu, J., Xu, G., Liu, Y., Liang, T., Zhang, X., Xiong, H. | https://link.springer.com/article/10.1007%2Fs11280-018-0596-8 | World Wide Web | Journal | 1 | 16 | \cite{DBLP:journals/www/YingWXLLZX19} | Successive POI recommendation | Yes | No | Yes(They discuss about latent space, I would vote for yes) | Yes | No | No | No | No | No | Yes(metric embedding) | No(They discuss about euclidean distance but in a latent space. I would say no) | No | No | No | Yes | Yes | Yes | No | No | No | Yes(80% training, 10% validation and 10% test) | No | No | No | No | Yes(For Foursquare, users with less than 10 Check-ins and POIs with less than 5 removed. For Gowalla, users visited less than 20 POIs removed and POIs visited by less than 15 users removed) | Yes(For fousquare they use New York, For Gowalla it is united states) | Ranking | Precision and Recall | Yes(Popular, BPR, FPMC, PRME) | Yes(Pop) | Yes(BPR) | No | Yes(10%) | Check-ins | No | Yes | Yes | No | No | No(They refer to Modeling user activity preference by leveraging user spatial temporal characteristics in lbsns 2015 for FS and Friendship and mobility: user movement in location-based social networks 2011 for Gowalla) | Yes | Post | No | Foursquare and Gowalla: Modeling user activity preference by leveraging user spatial temporal characteristics in lbsns 2015 for Foursquare and Friendship and mobility: user movement in location-based social networks 2011 for Gowalla | ||||||||||||||||||||
325 | 2019 | VCG | VCG: Exploiting visual contents and geographical influence for Point-of-Interest recommendation | Zhang, Z., Zou, C., Ding, R., Chen, Z. | Neurocomputing | Journal | 1,2 | 2 | \cite{DBLP:journals/ijon/ZhangZDC19} | No(Computes a similarity matrix but no neighbours) | Yes(part of the hybrid approach) | Yes(part of the hybrid approach) | No | No | Yes(Visual content + community) | Alternate Least squares | No | No | Yes | Yes | Yes(images, visual features) | No | No | No | Yes | No | No | No | No | No | Yes(80% locations for training, 20% to test) | No | No | Yes(for yelp, removed POIs visited by least 2 distinct users and remove users with less than 5 Check-ins. For Breadtrip, same for POIs and removed users with less than 8 Check-ins) | No | Ranking | Precision and Recall | No(UCF, ICF, NMF, WRMF, VBPR) | No | Yes(UCF, ICF) | No | No(5-fold cross validation for tune the parameters) | POIs | No | No | No | Yes | BreadTrip | No | Not complete (Check-ins not stated) | None | No | Yelp and Breadtrip: no further details | |||||||||||||||||||||||
326 | 2019 | RealTime-MF | Real-time event embedding for POI recommendation | Hao, P.-Y., Cheang, W.-H., Chiang, J.-H. | Neurocomputing | Journal | 1,2 | 10 | \cite{DBLP:journals/ijon/HaoCC19} | No | Yes(most important part) | No | Yes(previous step to MF) | No | No | No | No | Yes | No | No | Yes(Categories) | Yes(textual, keyword extraction) | No | Yes | Yes | No | No | No | No | Yes(80% training, 10% validation 10% test but not clear if it is CC random or temporal. Assume random CC) | No | No | No | Yes(extracted for a previous work) | Yes(New York) | Ranking | Recall, MRR | No(MF, MP, CDL, ConvMF, DCPR) | No | Yes(MF) | No | Yes(10% validation) | Check-ins | No | No | Yes ( to enrich the data) | No | No | No(They refer to GeoBurst: real-rime local event detection in geo-tagged tweet streams 2016) | Not complete (Only POIs stated) | None | No | Foursquare: They refer to GeoBurst: real-rime local event detection in geo-tagged tweet streams 2016 | |||||||||||||||||||||||
327 | 2019 | ST-RNet | ST-RNet: A Time-aware Point-of-interest Recommendation Method based on Neural Network | Gao, L., Li, Y., Li, R., Zhu, Z., Gu, X., Habimana, O. | https://ieeexplore.ieee.org/document/8852377 | IJCNN | Conference | 1 | 1 | \cite{DBLP:conf/ijcnn/GaoLLZGH19} | Time Aware | No | No | No | Yes | No | No | No | No | No(but in the DNN) | Yes | No | No | No | No | Yes | Yes | No | No | No | No | No | Yes(62.5% for training for each user, rest to test) | No | No | Yes(removed users and POIs with less than 5 Check-ins) | No. Although they refer to UTESE and I know they filter by city there (Singapore for Foursquare) | Ranking | Precision, Recall | No(LRT,UTE,UTE+SE,TAG-BPP,GTAG-BPP,UPT) | No | No | Yes(GTAG-BPP, UTESE) | No | POIs | No | No | Yes | No | No | No(They refer to Time- aware point-of-interest recommendation 2013) | Yes | Post | No | Foursquare: Time- aware point-of-interest recommendation 2011 | |||||||||||||||||||||
328 | 2019 | LC-G-P | An efficient location recommendation scheme based on clustering and data fusion | Cai, W., Wang, Y., Lv, R., Jin, Q. | Computers & Electrical Engineering | Journal | 1,2 | 4 | \cite{DBLP:journals/cee/CaiWLJ19} | Yes | No | No | No | Yes | Yes(geographical, preferences and location popularity) | No | Yes | No | Yes | Yes | No | No | No | No | Yes | No | No | No Information | No Information | No Information | No Information | No Information | No Information | Yes(remove users with less than 10 Check-ins and less than 30 social interactions and pois with less than 5 Check-ins) | No | Ranking | Precision, Recall | None(versions of the algorithm) | Yes(Pop) | No | Yes(version of the proposed algorithm with geogrphical) | No | No information | No | No | Yes | No | No | No | Yes | Prev and Post (i will indicate post) | No | Foursquare: no further information | |||||||||||||||||||||||
329 | 2019 | ADPR | ADPR: An Attention-based Deep Learning Point-of-Interest Recommendation Framework | Yin, J., Li, Y., Liu, Z., Xu, J., Xia, B., Li, Q. | https://ieeexplore.ieee.org/document/8852309 | IJCNN | Conference | 1 | 1 | \cite{DBLP:conf/ijcnn/YinLLXXL19} | No | No | Yes | Yes | No | No | Stochastic Gradient Descent | No | No(but in the DNN) | Yes | No | Yes | No | No | No | Yes | No | No | No | Yes(70% training, 30% test) | No | No | No | No | Yes(removed users and POIs with less than 5 Check-ins) | Yes(Manhattan) | Ranking | Precision, Recall and F1 measure | No(UCF+G, POI2Vec, GeoIE, versions of the algorithm) | No | No | Yes(UCF, GEoIE) | No | Check-ins | No | No | Yes | No | No | No(They refer to Attention-based recurrent neural network for location recommendation 2017) | Yes | Post | No | Foursquare: They refer to Attention-based recurrent neural network for location recommendation 2017 | ||||||||||||||||||||||
330 | 2019 | PRFPF | Personalized Point-of-Interest Recommendation on Ranking with Poisson Factorization | Su, Y., Li, X., Tang, W., Zha, D., Xiang, J., Gao, N. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8852462 | IJCNN | Conference | 1 | 1 | \cite{DBLP:conf/ijcnn/SuLTZXG19} | No | Yes(Poisson factorization) | Yes(they use BPR to optimize and Poisson factorization) | No | No | No | BPR | Yes | No | Yes | Yes | No | No | No | No | Yes | No | No | No | Yes(70% training, 20% test and 10% validation | No | No | No | No | Yes(removed users and POIs with less than 10 Check-ins) | No | Ranking | Precision, MAP, Recall and NDCG | No(Geosoca, igslr, BPR-MF, GeoBPR, GS2D, SG, BPR-KNN) | No | Yes(BPRMF) | Yes(GeoBPR, IGSLR) | Yes(10% validation) | Check-ins | No | No | Yes | Yes | No | No(They refer to Point-of-interest recommendations: Learning potential check-ins from friends, 2016 for Yelp and An experimental evaluation of point-of-interest recommendation in location-based social networks, 2017 for Foursquare) | Yes | Post | No | Yelp and Foursquare: Point-of-interest recommendations: Learning potential check-ins from friends 2016 for Foursquare and “An experimental evaluation of point-of-interest recommendation in location-based social networks 2017 for Yelp | ||||||||||||||||||||||
331 | 2019 | ---No-Acronym--They use "Flexible" | Flexible pOI recommendation based on user situation | Jang, S., Kim, J.-H., Nasridinov, A. | https://ieeexplore.ieee.org/document/8875378 | GreenCom/CPSCom | Conference | 1 | 0 | \cite{DBLP:conf/ithings/JangKN19} | No | No | No | No | Yes | No | No | No | No | Yes | No | No | No | Yes | No | Yes | No | No | No | No | No | Yes(80% training, 20% test) | No | No | Yes(removed users and POIs with less than 10 Check-ins) | Yes(New York) | Ranking | Accuracy and MAP | No baselines (apart from different components of the proposed method) | No | No | Yes(distance) | No | Check-ins | No | No | Yes | No | No | No | Yes | Post | No | Foursquare: no further information | ||||||||||||||||||||||
332 | 2019 | GT-HAN | A geographical-temporal awareness hierarchical attention network for next point-of-interest recommendation | Liu, T., Liao, J., Wu, Z., Wang, Y., Wang, J. | https://dl.acm.org/citation.cfm?doid=3323873.3325024 | ICMR | Conference | 1 | 5 | \cite{DBLP:conf/mir/LiuLWWW19} | Next POI recommendation | No | No | No(Only for optimization) | Yes | No | No | No | No | No(but in the DNN) | Yes | No | No | No | Yes(They say they ignore the sequential component but then state it. They exploit trajectories) | Yes | Yes | No | No | No | Yes(for each user first L Check-ins to train and rest to test) | No | No | No | No | Yes(remove users and POIs with less than 20 cehckins) | Yes(Foursquare USA, Gowalla worldwide) | Ranking | Accuracy and AUC | No(BPR, Geo-teaser, ATRank, Bi-LSTM, Bi-LSTM+Attention) | No | Yes(BPR) | No | No(5-fold validation to tune the parameters) | Check-ins | No | Yes | Yes | No | No | No(They refer to An experimental evaluation of point-of-interest recommendation in location- based social networks. 2017) | Yes | Post | No | Foursquare and Gowalla: They refer to An experimental evaluation of point-of-interest recommendation in location- based social networks 2017 | |||||||||||||||||||||
333 | 2019 | LORI | LORI: A Learning-to-Rank-Based Integration Method of Location Recommendation | Li, J., Liu, G., Yan, C., Jiang, C. | https://ieeexplore.ieee.org/document/8688654 | IEEE Transactions on Computational Social Systems | Journal | 1 | 4 | \cite{DBLP:journals/tcss/LiLYJ19} | Time Aware | Yes(RMCI) | No | Yes | No | No | Yes(it is learning to rank combining three components) | No | No | No | Yes | No | No | No | No | Yes | Yes | No | No | No | Yes(70% training and 30% test for each user ordered) | No | No | No | No | Yes(removed users and POIs with less than 6 Check-ins) | Yes(North America) | Ranking | Precision and Recall | No(UNIONgc,LLW,BPR) | No | No | Yes(RMGI) | No | Check-ins | No | Yes | No | No | No | No(They refer to Friendship and mobility: User movement in location-based social networks 2011) | Yes | Post | No | Gowalla: Friendship and mobility: User movement in location-based social networks 2011 | |||||||||||||||||||||
334 | 2019 | LSTM-S(Not sure. Stated as NA) | A Semantic sequential correlation based lstm model for next poi recommendation | Zhan, G., Xu, J., Huang, Z., Zhang, Q., Xu, M., Zheng, N. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8788754 | MDM | Conference | 1 | 0 | \cite{DBLP:conf/mdm/ZhanXHZX019} | Next POI recommendation | No | No | No | Yes | No | No | No | Yes | No | No | No | Yes | No | Yes | No | Yes | No | No | No | Yes(80% training, 20% test for each user) | No | No | No | No | Yes(removed users with less than 10 Check-ins and non-residential Check-ins) | Yes(USA) | Ranking | MRR,Recall | No(BPR, and variants of LSTM) | No | Yes(BPR) | No | No | Check-ins | No | No | Yes | No | No | No | Not complete (number of POIs not stated) | Post | No | Foursquare: Na- tiontelescope: Monitoring and visualizing large-scale collective behavior in lbsns 2015. Wrong stats for Foursquare, I think | |||||||||||||||||||||
335 | 2019 | PEU-RNN and UGSE-LR (UGSE-LR is already proposed so only analyze PEU-RNN) | On successive point-of-interest recommendation | Lu, Y.-S., Shih, W.-Y., Gau, H.-Y., Chung, K.-C., Huang, J.-L. | https://link.springer.com/article/10.1007%2Fs11280-018-0599-5 | World Wide Web | Journal | 1,3 | 8 | \cite{DBLP:journals/www/LuSGCH19} | Successive POI recommendation | No (affirmative for UGSE-LR) | No | No | Yes(for PEU-RNN) | No(They use a POI-POI graph but not social) | No (affirmative for UGSE-LR) | No | No | Yes(word2vec for PEU-RNN) | Yes(for UGSE-LR and for PEU-RNN they use the distance threshold) | No | No | No | Yes(for UGSE-LR and PEU-RNN) | No | Yes | No | No | No | No | Yes(70% training, 10% validation and 20%test) | No | No | No | Yes(removed POIs with less than 80 users cheked and users with less than 5 Check-ins) | No | Ranking | Precision and Recall | No(FPMC, FPMC-LR, POI2VEC) | No | No | No | Yes(10% validation) | Check-ins | No | Yes | No | No | Brightkite | Yes(http://snap.stanford.edu/data) | Yes | Post | No | Gowalla and Brighkite: http://snap.stanford. edu/data | |||||||||||||||||||||
336 | 2019 | MATI | Leveraging multi-aspect time-related influence in location recommendation | Hosseini, S., Yin, H., Zhou, X., Sadiq, S., Kangavari, M.R., Cheung, N.-M. | https://link.springer.com/article/10.1007%2Fs11280-018-0573-2 | World Wide Web | Journal | 1,3 | 11 | \cite{DBLP:journals/www/HosseiniYZSKC19} | No | Yes(They say it is a factor model) | Yes | No | No | No | Expectation Maximization | Yes | No | No | No | No | No | No | Yes | Yes | No | No | No | No | No | Yes(30% of the locations of every user to test) | No | No | No | No | Ranking | Precision, Recall, F1 | No(UBCF, USG, USGT, UBCTF, LRT) | No | Yes(UB) | Yes(USG) | No | POIs | No(They discuss about it) | No | Yes | No | No | Yes(https://snap.stanford.edu/data/loc-brightkite.html for brightkite and http://www.public.asu.edu/~hgao16/ for FOursquare although this one does not work) | Yes | Prev-No filtering | No | Foursquare and Brighkite: http://www.public.asu.edu/∼hgao16/ for FOursquare and https://snap.stanford.edu/data/loc-brightkite.html for Brighkite | ||||||||||||||||||||||
337 | 2019 | R2SIGTP | R2SIGTP: A novel real-time recommendation system with integration of geography and temporal preference for next point-of-interest | Jiao, X., Xiao, Y., Zheng, W., Wang, H., Jin, Y. | https://dl.acm.org/citation.cfm?doid=3308558.3314120 | WWW | Conference | 1 | 7 | \cite{DBLP:conf/www/JiaoXZWJ19} | Next POI recommendation | No | Yes | No | No | No | Yes(geographical and preference probability) | No | No | No | Yes | No | No | No | No | Yes | Yes | No | No | No | No | No | No | No | No | Evaluation for 200 users (no further information) | Yes(invalid Check-ins removed No further information) | Yes(New York) | Ranking | NDCG | No(only denoted as baseline. No further details provided) | No | No | No | No | No information | No | No | Yes | No | No | No | Not complete (number of users not stated) | None | No | Foursquare: no further information | ||||||||||||||||||||
338 | 2019 | DMGM-T or LDA-3. Not sure | A Geographical Behavior-Based Point-of-Interest Recommendation | Yu, X., Li, X., Li, J., Gai, K. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8818975 | HPSC-IDS | Conference | ??? | 1 | 0 | \cite{Yu2019} | Possible next POI | No | Yes(LDA based) | Yes(LDA based) | No | No | Yes(2 probabilities added) | No | Yes | No | Yes | No | Yes | No | No | Yes | Yes | No | No | No Information | No Information | No Information | No Information | No Information | No Information | Yes(users who have clicked more than 30 times) | Yes(New York and Los Angeles) | Ranking | Precision | No(R.S, LDA, MF) | No | Yes(MF, but not explained) | Yes(MGM) | Yes(10% validation) | No information | No | No | Yes | No | No | No(They refer to Location-based and preference- aware recommendation using sparse geo-social networking data 2012) | Not complete (number of Check-ins not stated) | None | No | Foursquare: They refer to Location-based and preference- aware recommendation using sparse geo-social networking data 2012 | ||||||||||||||||||||
339 | 2019 | PA-Seq2Seq | Context-aware attention-based data augmentation for POI recommendation | Li, Y., Luo, Y., Zhang, Z., Sadiq, S., Cui, P. | https://ieeexplore.ieee.org/document/8750927 | ICDE | Conference | 1 | 2 | \cite{DBLP:conf/icde/LiLZSC19} | Next POI recommendation | No | No | No | Yes | No | No | Gradient Descent | No | No(but in the DNN) | Yes(Distance in section III. A) | No | No | No | Yes | Yes | Yes | No | No | No | Yes(80% training for each user, 20 % test ordered also using validation) | No | No | No | No | No | No | Ranking | Accuracy | No(GRU, LSTM, RNN) | No | No | No | Yes(10% validation) | Check-ins | No | Yes | No | No | Brightkite | No | Yes | Prev-No filtering | No | Gowalla and Brighkite: no further information | |||||||||||||||||||||
340 | 2019 | HMM | Point-of-interest category recommendation based on group mobility modeling | Liu, X., Huang, X., Wang, Y., Zhang, L. | https://dl.acm.org/citation.cfm?id=3308670 | IUI | Conference | 1 | 1 | \cite{DBLP:conf/iui/LiuHWZ19} | Not sure if we should include this paper. Only 2 pages | No | No | Yes | No | No | No | No | Yes | No | No | No | Yes | No | Yes(Transition) | No | Yes | No | No | No Information | No Information | No Information | No Information | No Information | No Information | No | No | Ranking | Accuracy | No(MCF, PMF, HITS) | No | No(PMF) | No | No | No information | No | No | No | No | No | Yes | Prev-No filtering | No | Wechat: no further information | ||||||||||||||||||||||
341 | 2019 | CPC | Content-aware point-of-interest recommendation based on convolutional neural network | Xing, S., Liu, F., Wang, Q., Zhao, X., Li, T. | https://link.springer.com/article/10.1007%2Fs10489-018-1276-1 | Applied Intelligence | Journal | 1 | 10 | \cite{DBLP:journals/apin/XingLWZL19} | No | Yes(MF with a neural network) | Yes | Yes | No | No | Gradient Descent | No | Yes(word vector embedding) | Yes | No | No(They only work with restaurants) | Yes(reviews) | No | No | Yes | No | No | No | No | Yes(80% as training, 10% validation, 10% test) | No | No | No | Yes(only data for food) | Yes(New York and Los Angeles) | Ranking | Precision and Recall | No(UCF, PMF, LCARS, SELR, STLR, CAPRF, VPOI) | No | Yes(UCF) | No | Yes(10% validation) | Check-ins | Yes(no training data for the users or items) | No | Yes | No | No | No(They follow the same crawling strategy from to Exploring social-historical ties on locationbased social networks 2012 | Yes | Post | No | Foursquare: They follow the same crawling strategy from Exploring social-historical ties on locationbased social networks 2012 | ||||||||||||||||||||||
342 | 2019 | NRLRS | Exploring iot location information to perform point of interest recommendation engine: Traveling to a new geographical region | Yang, X., Zimba, B., Qiao, T., Gao, K., Chen, X. | https://www.mdpi.com/1424-8220/19/5/992 | Sensors | Journal | 1 | 1 | \cite{DBLP:journals/sensors/0003ZQGC19} | No | Yes | Yes | No | Yes | No | No. Weighted Category Hierarchical. I would say no | Gradient Descent | No | No | Yes | Yes | Yes | Yes(text reviews) | No | No | Yes | No | No | No | No | Yes(normal standard) | No | No | No | No | Yes(Las Vegas, Phoenix) | Error | MAE, RMSE | No(UB-KNN, IB-KNN, UC, CKNN, CVD++, HFT) | No | Yes(uknn, iknn) | No | No | POIs(Equivalent) | No | No | No | Yes | No | No | Yes | Prev-No filtering | No | Yelp: no further information | |||||||||||||||||||||
343 | 2019 | MLR | A two-step personalized location recommendation based on multi-objective immune algorithm | Geng, B., Jiao, L., Gong, M., Li, L., Wu, Y. | Information Sciences | Journal | 1,2 | 15 | \cite{DBLP:journals/isci/GengJGLW19} | Yes | Yes(Social Collaborative Filtering) | No | Yes(KDE) | No | No | No | Yes(Genetic algorithm) | No | No | No | Yes | Yes | No | No | No | No | Yes | No | No | No | No | Yes(80% training, 20% test) | No | No | No | No | Yes(Seattle, New York and Austin) | Ranking | Precision, Recall and F1 measure | No(CF, SCF, NBI, KDE, MLR, NSGALR) | No | Yes(CF) | Yes(KDE) | No | POIs | No | Yes | No | No | Brightkite | Yes(http://snap.stanford.edu/data/loc-gowalla.html for Gowalla and http://snap.stanford.edu/data/loc-brightkite.html for Brightkite and they also refer They refer to iGSLR: personalized geo-social location recommendation: a kernel density estimation approach 2013) | No | None | No | Gowalla and Brighkite: They refer to iGSLR: personalized geo-social location recommendation: a kernel density estimation approach 2013 and also http://snap.stanford.edu/data/loc-gowalla.html for Gowalla and http://snap.stanford.edu/data/loc-brightkite.html for Brighkite | |||||||||||||||||||||
344 | 2019 | UGR | Personalized POI recommendation based on check-in data and geographical-regional influence | Song, C., Wen, J., Li, S. | https://dl.acm.org/citation.cfm?id=3311034 | ICMLSC | Conference | ??? | 1 | 1 | \cite{Song2019} | Yes(part of the hybrid approach) | No | No(power law distribution) | No | No | Yes(CF + Geographical influence) | No | Yes(Density-based spatial clustering and Noise (DBSCAN)) | No | Yes | No | No | No | No | No | Yes | No | No | No | No | Yes(75% training, 25% test) | No | No | No | Yes(removed POIs and users with less than 5 Check-ins) | No | Ranking | Precision and Recall | No(U, G, MGM, USG) | No | Yes(U) | Yes(USG) | No | Check-ins | No | Yes | No | No | No | No | Yes | Post | No | Gowalla: no further details | |||||||||||||||||||||
345 | 2019 | U-CF-Memory | Discovering memory-based preferences for POI recommendation in location-based social networks | Gan, M., Gao, L. | https://www.mdpi.com/2220-9964/8/6/279/htm | International Journal of Geo-Information | Journal | 1 | 4 | \cite{DBLP:journals/ijgi/GanG19} | Yes | No | No | No | No | No | No | No | No | No | No | No | No | No | Yes | Yes | No | No | Yes(80% training, 20% test) | No | No | No | No | No | Yes(removed users with less than 10cehckins and POIs with less than 10 Check-ins) | Yes(New York City) | Ranking | Precision, Recall and F-Measure | No(U-CF) | No | Yes(U-CF) | No | No(10-fold validation) | Check-ins | No | No | Yes | No | No | No | Yes | Post | No | Foursquare: no further information | ||||||||||||||||||||||
346 | 2019 | AGS-MF | Modeling heterogeneous influences for point-of-interest recommendation in location-based social networks | Guo, Q., Sun, Z., Zhang, J., Theng, Y.-L. | https://link.springer.com/chapter/10.1007%2F978-3-030-19274-7_6 | ICWE | Conference | 1 | 0 | \cite{DBLP:conf/icwe/GuoSZT19} | No | Yes | No | No | Yes | No | Stochastic Gradient Descent | No | No | Yes | Yes | No(They discuss about categories but not sure how. I would say No because the aspects are obtained from the reviews) | Yes(reviews) | No | No | Yes | No | No | No | Yes(80% training for each user, 20 % test ordered) | No | No | No | No | No | Yes(Phoenix, Las Vegas and Charlotte) | Ranking | Precision, Recall and MAP | No(UCF, ICF, MF, SRMF, LFBCA, GeoMF, GeoSOca, TriRank) | No | Yes(UCF) | Yes(GeoMF) | No | Check-ins | No | No | No | Yes | No | Yes(https://www.yelp.com/dataset/challenge) | Yes | Prev-No filtering | No | Yelp: https://www.yelp.com/dataset challenge | ||||||||||||||||||||||
347 | 2019 | ---No-Acronym-- | Discovering travel community for POI recommendation on location-based social networks | Tang, L., Cai, D., Duan, Z., Ma, J., Han, M., Wang, H. | https://www.hindawi.com/journals/complexity/2019/8503962/ | Complexity | Journal | 1 | 8 | \cite{DBLP:journals/complexity/TangCDMHW19} | Yes | No | Yes | No | No | Yes(but it combines social and probabilistic) | No | No | No | No | Yes | Yes(In algorithm 1 the say about categories) | No | Yes(Markov) | No | Yes | No | No | No Information | No Information | No Information | No Information | No Information | No Information | Yes(false Check-ins removed -e.g. only one Check-in of a user in a day) | Yes(New York) | Ranking | Accuracy(Precision) and Recall | No(CFbased, Commendbased and TCBased) | No | Yes(CF) | No | No | No information | No | Yes | Yes | No | No | No(They can provide the data if requested) | Yes | Post | No | Foursquare and Gowalla: no further details provided | ||||||||||||||||||||||
348 | 2019 | APRA-SA | An adaptive point-of-interest recommendation method for location-based social networks based on user activity and spatial features | Si, Y., Zhang, F., Liu, W. | Knowledge Based Systems | Journal | 1,2 | 16 | \cite{DBLP:journals/kbs/SiZL19} | Yes | No | No | Yes(gaussian kernels and power law) | No | No | Yes(time popularity and distance based) | Clustering | No | Yes(user activity clustering) | No | Yes | No | No | No | No | Yes | Yes | No | No | No | No | No | Yes(for each user 84% of the Check-ins as training data and rest to test) | No | No | No | No | Error/Ranking | Precision, Recall, F1, MAE and RMSE | No(SB, UCF, SK, UTE+SE, GTAG-BP, GT-BNMF, CTF-ARA) | No | Yes(UBCF) | Yes(SB) | No | Check-ins | No | Yes | Yes | No | No | No(They refer to Time-aware point-of-interest recommendation 2013) | Yes | Prev-No filtering | No | Foursquare and Gowalla: They refer to Time-aware point-of-interest recommendation 2013 | |||||||||||||||||||||
349 | 2019 | Geo-SRank | Network Embedding-Aware Point-of-Interest Recommendation in Location-Based Social Networks | Guo, L., Jiang, H., Liu, X., Xing, C. | https://www.hindawi.com/journals/complexity/2019/3574194/ | Complexity | Journal | 1 | 2 | \cite{DBLP:journals/complexity/GuoJLX19} | No | Yes | Yes(similar to BPR to optimize) | Yes(pretrained embeddings. Factorization is Yes) | No | No | No | No | Yes(node2vec) | Yes | Yes | No | No | No | No | Yes | No | No | No | No | No | Yes(70% for training, 10%validation, 20% test. Repeated 5 times). In cold start evaluation methodology changes | No | No | Yes(For Gowalla, users with Check-ins fewer than 15 Check-ins removed and POIs with less than 10 visitors) | No | Ranking | Precision, Recall, MAP | Yes(MostPopular, WRMF, GeoMF, BPRMF, RankGeoFM, IrenMF) | Yes(POp) | Yes(BPRMF) | Yes(IREMF) | Yes(10% validation) | POIs | Yes(no training data for the users or items) | Yes | No | Yes | No | Yes(https://snap.stanford.edu/data/loc- gowalla.html) for Gowalla and An experimental evaluation of point-of-interest recommendation in location-based social networks 2017 for Yelp) | Yes | Post (contradictory data in the statistics) | No | Gowalla and Yelp: They refer to An experi- mental evaluation of point-of-interest recommendation in location-based social networks 2017 | ||||||||||||||||||||||
350 | 2019 | SG- NeuRec | Deep Neural Model for Point-of-Interest Recommendation Fused with Graph Embedding Representation | Zhu, J., Guo, X. | https://link.springer.com/chapter/10.1007%2F978-3-030-23597-0_40 | WASA | Conference | 1 | 0 | \cite{DBLP:conf/wasa/ZhuG19} | No | No | No | Yes | No | No | No | No | Yes(graph embedding) | Yes | Yes | No | No | No | Yes | Yes | No | No | No | Yes(For each user, we use their recent interactions as test and select 100 unvisited POIs that are not accesed by the user to check. Leave one out) | No | No | No | No | No | No | Ranking | HitRate and MRR | No(BPR-MF, SoRec, GE, NeuMF, PACE) | No | Yes(BPR) | GE | No | Check-ins | No | Yes | No | Yes | No | No | Yes | Prev-No filtering | No | Gowalla and Yelp: no further details | ||||||||||||||||||||||
351 | 2019 | SSSER | SSSER: Spatiotemporal sequential and social embedding rank for successive point-of-interest recommendation | Xu, Y., Li, X., Li, J., Wang, C., Gao, R., Yu, Y. | https://ieeexplore.ieee.org/document/8886571 | IEEE Access | Journal | 1 | 2 | \cite{DBLP:journals/access/XuLLWGY19} | Successive POI recommendation | No | No | Yes(to optimize) | Yes(most important part) | No | SGD and BPR | No | No(but in the DNN) | Yes | Yes | No | No | Yes | No | Yes | No | No | No | No | Yes(70% of the data for training, 10% as validation and 20% as test) | No | No | No | Yes(For Foursquare, users with less than 10 Check-ins and POIs with less than 15 removed. For GOwalla, users with less than 10 Check-ins removed) | No | Ranking | Precision, Recall, MAP and NDCG | No(FPMC, Fossil, HRNN, PRME, HRNN) | No | No | Yes(IGSLR, USG) | Yes(10% validation) | Check-ins | No | Yes | Yes | No | No | No(They refer to gSCorr: Modeling geo-social cor- relations for new check-ins on location-based social networks 2012 for Foursquare and Friendship and mobility: User movement in location-based social networks 2011 for Gowalla) | Yes | Post | No | Foursquare and Gowalla: gSCorr: Modeling geo-social cor- relations for new check-ins on location-based social networks 2012 for Foursquare and Friendship and mobility: User movement in location-based social networks 2011 for Gowalla | ||||||||||||||||||||||
352 | 2019 | BLR | Behavior-based location recommendation on location-based social networks | Rahimi, S.M., Far, B., Wang, X. | https://link.springer.com/article/10.1007/s10707-019-00360-3 | GeoInformatica | Journal | ??? | 1 | 5 | \cite{DBLP:journals/geoinformatica/RahimiFW20} | Although it seems the same as \cite{DBLP:conf/pakdd/RahimiWF17}, their proposal is different. Although the bib is 2020, in reality is 2019 | No | No | Yes(2 probabilities combined) | No | No | Yes(2 probabilities combined) | No | Yes(Density-based spatial clustering and Noise (DBSCAN)) | No | Yes | No | Yes | No | Yes | Yes | Yes | No | No | No | No | No | Yes(1 random Check-ins per user to test) | No | No | No | No | Ranking | Precision, Recall, Behaviour precision and Spatial Precision | No(GeoMF, GeoMF++,MLR,USPB,PMM,USG) | No | No | Yes(GeoMF, USG) | No | Check-ins | Yes(users with less than 5 chekcins in hte dataset) | Yes | No | No | No | Yes(They refer to A study ofrecommending locations on location-based social network by collaborative filtering 2012 for Gowalla) | Yes | Prev-No filtering | No | Gowalla: They refer to A study of recommending locations on location-based social network by collaborative filterin 2012 | ||||||||||||||||||||
353 | 2019 | ASTEN | An attentive spatio-temporal neural model for successive point of interest recommendation | Doan, K.D., Yang, G., Reddy, C.K. | https://link.springer.com/chapter/10.1007%2F978-3-030-16142-2_27 | PAKDD | Conference | 1 | 5 | \cite{DBLP:conf/pakdd/DoanYR19} | Successive POI recommendation | No | No | No | Yes | No | No | No | No | No | Yes | No | No | No | Yes | Yes | Yes | No | No | No | No | No | No | Yes (5 fold cross validation CC) | No | Yes(filter out POIs and users with less than 10 visits) | Yes(Foursquare, Europe and USA) | Ranking | Recall, F1 and AUC | No(TOP, MC, LRTL, FPMC, PRME, RNN, ST-RNN) | Yes(POp) | No | Yes(ST-RNN) | No(cross-validation) | Check-ins | No | Yes | Yes | No | No | No | Yes | Post | No | Foursquare and Gowalla: no further details | |||||||||||||||||||||
354 | 2019 | HeteGeoRanRec | Modeling user contextual behavior semantics with geographical influence for point-of-interest recommendation | Yu, D., Xu, K., Wang, D. | http://ksiresearchorg.ipage.com/seke/seke19paper/seke19paper_178.pdf | SEKE | Conference | 1 | 1 | \cite{DBLP:conf/seke/YuXW19} | No | Yes | Yes | No | No | No | Alternate Least squares | No | No | Yes | Yes | Yes | No | No | No | Yes | No | No | No | Yes(80% training, 20% test for each user ordered by timestamps) | No | No | No | No | No | No | Ranking | Precision and Recall | No(BPRMF, WRMF, USG, GMF, RankGeoMF, ASMF) | No | Yes(BPR) | Yes(USG) | No | Check-ins (But they aggregate them) | No | No | Yes | No | No | No(They refer to Point-of-interest recommendations: Learning potential check-ins from friends 2016) | Yes | Prev-No filtering | Yes(https://github.com/Skyexu/HeteGeoRankRec) | Yes | Yes | Yes | Foursquare: https://dropbox.com/s/pa1mni3h8qdkdby/Foursquare.zip?dl=0 and refer to Point-of-interest recommendations: Learning potential check-ins from friends 2016 | |||||||||||||||||||
355 | 2019 | STA | Spatiotemporal representation learning for translation-based POI recommendation | Qian, T., Liu, B., Nguyen, Q.V.H., Yin, H. | https://dl.acm.org/citation.cfm?doid=3306215.3295499 | TOIS | Journal | 1 | 37 | \cite{DBLP:journals/tois/QianLNY19} | Yes | Yes(cold start) | Yes(embedding con factorization) | No | No | No | No | No | No | Yes | Yes | No | No | No | No | Yes | Yes | No(although they claim to use online recommendation) | No | No | Yes(80% training, 20% test for each user ordered by timestamps, also 10% of validation) | No | No | No | No | No | No | Ranking | Recall and NDCG | No(USG, LRT, GeoMF, RankGeoMF, Ge, TransRec, LORE, MGMOPMF) | No | No | Yes(USG, GeoMF) | Yes(10% of the last Check-ins) | Check-ins | Yes | Yes | Yes | No | No | Yes(https://sites.google.com/site/dbhongzhi/) | Yes | Prev-No filtering | No | Foursquare and Gowalla: https:/sites.google.com/site/dbhongzhi ACM | |||||||||||||||||||||
356 | 2019 | WBPR-DST | Personalized ranking point of interest recommendation based on spatial-temporal distance metric in LBSns | Su, C., Li, H., Xie, X. | https://dl.acm.org/citation.cfm?id=3316715 | ICSCA | Conference | ??? | 1 | 0 | \cite{Chang2019} | No | No | Yes(BPR in a proposed metric) | No | No | No | Yes(Distance metric with BPR) | BPR | No | No | Yes | No | No | No | No | Yes | Yes | No | No | No | No | Yes(80% training, 20% test but not clear if random or temporal CC) | No | No | No | Yes(remove users with less than 10 Check-ins and remove POIs checked less than 30 times) | No | Ranking | Precision, HitRate and NDCG | No(Stellar, BPRMF) | No | Yes(BPR) | Yes(STELLAR) | No | Check-ins (But they aggregate them) | No | Yes | No | No | Brightkite | No | Not complete (number of Check-ins not stated) | None | No | Brighkite and Gowalla: no further details | ||||||||||||||||||||
357 | 2019 | RBMNMF | A deep learning model based on sparse matrix for point-of-interest recommendation | Zeng, J., Tang, H., Li, Y., He, X. | http://ksiresearchorg.ipage.com/seke/seke19paper/seke19paper_156.pdf | SEKE | Conference | 1 | 2 | \cite{DBLP:conf/seke/ZengTLH19} | No | Yes | No | Yes | No | Yes | No | No | No | No | No | No | No | No | No | Yes | No | No | No | No | No | Yes(25% to test, 12.5% to valdiation, rest to train for every user) | No | No | Yes(remove users and items with less than 5 check-ins) | Yes(Singapore) | Ranking | Precision and Recall | No(UB-KNN, RBM, NMF, RBMNMF) | No | Yes(CF, BPR) | No | Yes | POIs | No | No | Yes | No | No | Yes(same one as used in Time-aware point-of-interest recommendation (SIGIR 2013) | Yes | Post | No | Foursquare: Time-aware point-of-interest recommendation 2013 | ||||||||||||||||||||||
358 | 2019 | ---No-Acronym-- | Exploring Spatial and Mobility Pattern's Effects for Collaborative Point-of-Interest Recommendation | Jiao, X., Xiao, Y., Zheng, W., Xu, L., Wu, H. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8889741 | IEEE Access | Journal | ??? | 1 | 6 | \cite{Jiao2019} | Yes | No | No | No | No | Yes(Eq.15. For me is a yes) | No | Yes(MeanShift) | No | Yes | No | Yes(They discuss about categories Eq 11) | No | Yes | Yes | Yes | No | No | Yes(last two months to test, rest to train) | No | No | No | No | No | Yes(users with less than 3 POIs removed) | Yes(New York, Tokyo) | Ranking | Precision, Recall, F1 | No | No | No | Yes(USG, ASMF, Geo-PFM, GeoSoCa, CoRe) | No | Check-ins (no info provided but pretty sure check-ins) | No | Yes | Yes | No | No | Yes(https://sites.google.com/site/yangdingqi/home/foursquare-dataset for Foursquare and http://snap.stanford.edu/data/loc-Gowalla.html for Gowalla) | Yes | Post | No | Fourquare and Gowalla: https://sites.google.com/site/yangdingqi/home/foursquare-dataset for Foursquare and http://snap.stanford.edu/data/loc-Gowalla.html for Gowalla | |||||||||||||||||||||
359 | 2019 | BPRSoReg | Social regularisation in a BPR-based venue recommendation system | Liu, S., Ounis, I., Macdonald, C. | http://ceur-ws.org/Vol-2537/paper-04.pdf | FDIA@ESSIR | Conference | 1 | 0 | \cite{DBLP:conf/fdia/LiuOM19} | No | Yes(It is a MF with BPR optimization) | Yes(BPR with social regularization) | No | No | No | BPR | No | No | No | Yes(friends) | No | No | No | No | Yes | No | No | No | No | Yes(80% training, 20% test) | No | No | No | No | No | Ranking | MRR | No | No | Yes(BPR) | No | No | Check-ins | Yes | No | No | Yes | No | No | Yes | Prev-No filtering | No(They provide the spotlight repository) | Yelp: no url provided | ||||||||||||||||||||||
360 | 2019 | STGN | Where to go next: A spatio-temporal gated network for next POI recommendation | Zhao, P., Zhu, H., Liu, Y., Xu, J., Li, Z., Zhuang, F., Sheng, V.S., Zhou, X. | https://ojs.aaai.org//index.php/AAAI/article/view/4537 | AAI | Conference | 1 | 45 | \cite{DBLP:conf/aaai/ZhaoZLXLZSZ19} | Next-poi recommendation | Yes | No | No | No | Yes | No | No | SGD (Stochastic Gradient Descent) | No | No | Yes(differences in distance) | No | No | No | Yes(LSTMs) | Yes(differences in time) | Yes | No | No | No | Yes(70% training, rest to test for each user) | No | No | No | No | Yes(users and POIs with less than 10 check-ins removed) | Yes(California and Singapore) | Ranking | Accuracy and MAP | No | No | No | Yes | No | Check-ins | Yes | Yes | Yes | No | Brightkite | Yes(only for gowalla and brightkite: http://snap.stanford.edu/data/loc-gowalla.html, http://snap.stanford.edu/data/loc-brightkite.html) | Yes | Post(not clear although the stats do not match of the original sources, so I state as post) | No | Foursquare: no further information. Gowalla and Brightkite http://snap.stanford.edu/data/loc-gowalla.html, http://snap.stanford.edu/data/loc-brightkite.html | ||||||||||||||||||||
361 | 2019 | GFP-LORE | A multi-element hybrid location recommendation algorithm for location based social networks | Yue-Qiang, R., Ze, W., Xiao-Na, S., Shi-Min, S. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8764552 | IEEE Access | Journal | 1 | 3 | \cite{DBLP:journals/access/Yue-QiangZXS19} | No | No | Yes(power-law) | No | No | Yes(Popularity + KDE + social + sequential) | No | No | No | Yes(power-law) | Yes(friends) | No | No | Yes(for visiting next-POI) | No | Yes | No | No | Yes(half of the check-ins to training, rest to test) | No | No | No | No | No | Yes(no further details) | No | Ranking | Precision and Recall | No(FPMC, AMC, GS2D, LORE) | No | No | Yes(GS2D, LORE) | No | Check-ins | No | Yes | No | No | No | No | Yes | Post | No | Gowalla: They refer to Friendship and mobility: User movement in location-based social networks 2011 | ||||||||||||||||||||||
362 | 2019 | TTR | Loc2Vec-Based Cluster-Level Transition Behavior Mining for Successive POI Recommendation | Wen, Y., Zhang, J., Zeng, Q., Chen, X., Zhang, F. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8776588 | IEEE Access | Journal | 1 | 1 | \cite{DBLP:journals/access/WenZZCZ19} | Successive POI recommendation | No | Yes | Yes | No | No | Yes(MF with geographical and fusing a model) | No | No | Yes(Word2Vec, Loc2Vec) | Yes(distance of the POIs) | No | No | No | Yes(Transition POIs) | No | Yes | No | No | No | No | Yes(not explicitly stated) | No | No | No | No | No | Ranking | Precision and Recall | No(FPMC, FMC, LORE, FPMC_LR) | No | No | Yes(LORE, FPMC_LR) | No | Check-ins | No | Yes | No | No | Brightkite | No | Yes | Prev-No filtering | No | Gowalla and Brightkite. No further details | |||||||||||||||||||||
363 | 2019 | DPGSR-PR | Deep potential geo-social relationship mining for point-of-interest recommendation | Pan, Z., Cui, L., Wu, X., Zhang, Z., Li, X., Chen, G. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8768295 | IEEE Access | Journal | 1 | 3 | \cite{DBLP:journals/access/PanCWZLC19} | No | Yes | Yes(KDE) | No | Yes(Random Walk) | No(All the model is a MF) | SGD (Stochastic Gradient Descent) | No | No | Yes(KDE with distances) | Yes(friends) | No | No | No | No | Yes | No | No | No | No | Yes(80% training, 20% test) | No | No | No | No | No | Ranking | Precision, Recall and NDCG | No(MF, PMF, GeoCF, CoRe) | No | Yes(MF) | Yes(GeoCF, CoRe) | No | Check-ins | No | Yes | Yes | No | No | No | Yes | Prev-No filtering | No | Gowalla. They refer to Fused matr ix factorization with geographical and social influence in location-based social networks 2012 Foursquare They refer to gSCorr: Modeling geo-social correlations for new check-ins on location-based social networks 2011 | ||||||||||||||||||||||
364 | 2019 | SGBA | Next POI Recommendation via Graph Embedding Representation from H-Deepwalk on Hybrid Network | Yang, K., Zhu, J. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8915810 | IEEE Access | Journal | 1 | 1 | \cite{DBLP:journals/access/YangZ19c} | Next-POI recommendation | No | No | No | Yes(LSTM) | Yes(DeepWalk) | Yes(SPTL + LTPL) | No | No | Yes(skip-gram, graph embedding) | Yes | Yes(friends) | No | No | Yes | No | Yes | No | No | No Information | No Information | No Information | No Information | No Information | No Information | No | No | Ranking | AUC | No(BPR, LSTM-Rec, ST-RNN, AT-Rank) | No | Yes(BPR) | Yes(ST-RNN) | No | Prev-no filtering | Yes | Yes | No | No | Brightkite | No | Yes | No information | No | Gowalla and Brightkite. THey say to obtain from Standfor university | |||||||||||||||||||||
365 | 2019 | ---No-Acronym-- | Striking the Balance between Novelty and Accuracy in Location-Based Recommendation System | Agrawal, V., Sahu, S., Oommen, S., Mohana Reddy, G.R. | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8960224 | Innovations in Power and Advanced Computing Technologies | Conference | ??? | 1 | 0 | \cite{Agrawal20019} | No | No | No | No | Yes | No | No | No | No | No | Yes(friends) | No | No | No | No | Yes | No | No | No | No | Yes(70% training, 30% test) | No | No | No | Yes(No further details) | No | Ranking | Precision, Recall | No(NMF and SR) | No | Yes(NMF) | No | No | Check-ins | No | No | Yes | No | No | No | Yes | Prev | No | Foursquare. No further details | |||||||||||||||||||||
366 | 2019 | ---No-Acronym-- | POI recommendation based on heterogeneous graph embedding | Mighan, S.N., Kahani, M., Pourgholamali, F. | https://ieeexplore.ieee.org/document/8964762 | International eConference on Computer and Knowledge Engineering (ICCKE) | Conference | ??? | 1 | 0 | \cite{Mighan2019} | No | No | No | Yes(skip-gram) | No | No | SGD (Stochastic Gradient Descent) | No | Yes(graph embedding) | No | No | No(They discuss about categories but not sure how they use them) | No | Yes(successive) | Yes(temporal nodes) | Yes | No | No | No | Yes(80% training, rest to test for each user) | No | No | No | No | No. Not sure if they remove cold start users but it seems not | No | Ranking | Precision, Recall, F1 | No(GE, PACE) | No | No | Yes(PACE, GE) | No | Check-ins | No. Not clear | No | Yes | No | No | No | Yes | Not clear if they remove cold start | No | Foursquare: They refer to Exploiting context graph attention for poi recommendation in location-based social networks. 2018 and Joint modeling of user check-in behaviors for point-of-interest recommendation 2015 | |||||||||||||||||||||
367 | 2019 | DLM | Personalized Recommendation Method of POI Based on Deep Neural Network | Gao, Y., Duan, Z., Shi, W., Feng, J., Chiang, Y.-Y. | https://ieeexplore.ieee.org/document/8963449 | International Conference on Behavioral, Economic and Socio-Cultural Computing, BESC | Conference | 1 | 1 | \cite{DBLP:conf/besc/GaoDS0C19} | No | Yes(MF) | Yes(LDA) | Yes(MF+ LDA sent to neural network) | No | No | Gradient Descent | No | Yes(feature embedding) | Yes | No | Yes(Feature embedding, they use the POI category) | No | No | No | Yes | No | No | No | No | No | Yes(80% training, 20%test) | No | No | Yes(users and POIs with less than 5 check-ins) | Yes(Beijing) | Ranking | Precision, Recall | No(UCF, PMF, LCARS, RankGeo-FM, SGFM) | No | Yes(UCF) | Yes(Rank-GeoFM, LCARS) | No | Check-ins | No | No | Yes | No | No | No | Not complete | Prev | No | Foursquare: no further details | ||||||||||||||||||||||
368 | 2019 | ---No-Acronym-- | Point of interest recommendation by exploiting geographical weighted center and categorical preference | Mo, F., Yamana, H. | https://ieeexplore.ieee.org/document/8955628 | IEEE International Conference on Data Mining Workshops | Conference | 1 | 0 | \cite{DBLP:conf/icdm/MoY19} | Yes | Yes(MF in the categories) | No | No | No | Yes(Geographical + UB, + MF...) | No | No | No | Yes | Yes | Yes | No | No | No | Yes | No | No | No | No | No | Yes(70% training, rest to test) | No | No | Yes(users and POIs with less than 5 check-ins) | Yes(London and Brooklying and Queens) | Ranking | Precision, Recall and F1 | No(UPS, SGFM) | No | No | Yes(UPS, SGFM) | No | POIs | No | No | No | No | Weeplaces | No | Yes | Post | No | Weeplaces: They refer to Personalized Point- of-Interest Recommendation by Mining Users’ Preference Transition 2013 | ||||||||||||||||||||||
369 | 2019 | NCFT | POI recommendation based on first-order collaborative filtering tree | Zhu, J., Ma, S., Li, J. | https://ieeexplore.ieee.org/document/9066125 | International Conference on Mobile Ad-hoc and Sensor Networks | Conference | 1 | 0 | \cite{DBLP:conf/msn/ZhuML19} | No | No | No | Yes | No | No | No | No | Yes(user, POI embeddings) | Yes | Yes | Yes(tags) | No | Yes | Yes | Yes | No | No | No | Yes(last POI of each user to test) | No | No | No | No | Yes(users with less than 15 check-ins and POIs with less than 10 interactions) | No | Ranking | AUC | No(GRU, ATrank) | No | No | Yes(GRU with distance features) | No | No | No | Yes | No | Yes | No | Yes(http://spatialkeyword.sce.ntu.edu.sg/eval-vldb17/data/Gowalla.zip, http://spatialkeyword.sce.ntu.edu.sg/eval-vldb17/data/Yelp.zip) | Yes | Post | No | Gowalla and Yelp. They refer to Friendship and Mobility: User Movement In Location-Based Social Networks 2011 and An experimental evaluation of point-of-interest recommendation in location-based social networks 2017 | ||||||||||||||||||||||
370 | 2019 | LGLMF | LGLMF: Local Geographical Based Logistic Matrix Factorization Model for POI Recommendation | Rahmani, H.A., Aliannejadi, M., Ahmadian, S., Baratchi, M., Afsharchi, M., Crestani, F. | https://link.springer.com/chapter/10.1007/978-3-030-42835-8_7 | Asia Information Retrieval Symposium | Conference | 1 | 3 | \cite{DBLP:conf/airs/RahmaniAABAC19} | No | Yes | No | No | No | Yes(MF+other semi-probabilistic model) | No | No | No | Yes | No | No | No | No | No | Yes | No | No | No | Yes(70% ancient check-ins for each user to test, 10% validation and rest to test) | No | No | No | No | Yes(users with less than 15 check-ins and POIs with less than 15 interactions for Gowalla, and users and POIs with less than 10 interactions for Foursquare) | No | Ranking | Precision, Recall and NDCG | No | No | No | Yes(PFMMGM, LRT, IGSLR, L-WMF) | Yes | Check-ins | No | Yes | Yes | No | No | Yes(http://spatialkeyword.sce.ntu.edu.sg/eval-vldb17/) | Yes | Post | No | Gowalla and Foursquare: They refer to An experimental evaluation of point- of-interest recommendation in location-based social networks | ||||||||||||||||||||||
371 | 2020 | HRPR | POI Recommendation Based on Heterogeneous Network | Wen, Y., Zhang, J., Chen, G., Chen, X., Chen, M. | https://link.springer.com/chapter/10.1007/978-981-13-9409-6_217 | International Conference in Communications, Signal Processing, and Systems | Conference | 1 | 0 | \cite{DBLP:conf/csps/WenZCCC19} | No | No | Yes | Yes | Yes(random-walk) | No | No | No | No | Yes(although not especifically state) | No | No | No | Yes(transition probability) | No | Yes | No | No | No | No | Yes(different ratios) | No | No | No | No | Yes(USA) | Error | MAE/RMSE | No | No | No | No | No | POIs | Yes | No | No | Yelp | No | No | Yes | Prev-No filtering | No | Yelp: no further details | ||||||||||||||||||||||
372 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
373 | 2020 | A recommendation algorithm for point of interest using time-based collaborative filtering | Zeng, J., He, X., Li, F., Wu, Y. | https://ideas.repec.org/a/ids/ijitma/v19y2020i4p347-357.html | 1 | 0 | \cite{DBLP:journals/ijitm/ZengHLW20} | IGNORE. pdf not found | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
374 | 2020 | Exploiting user check-in data for geo-friend recommendations in location-based social networks | Liu, S., Zhang, K. | 1 | 0 | \cite{DBLP:journals/ijmcmc/LiuZ20} | IGNORE. pdf not found | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
375 | 2020 | On a method for location and mobility analytics using location-based services: a case study of retail store recommendation | Chen, Y.-M., Chen, T.-Y., Chen, L.-C. | 1 | 0 | IGNORE. pdf not found | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
376 | 2020 | An intelligent location recommender system utilising multi-agent induced cognitive behavioural model | Ravi, L., Devarajan, M., V, V., Sangaiah, A.K., Wang, L., A, S., Subramaniyaswamy, V. | Enterprise Information Systems | Journal | ??? | 1 | 0 | IGNORE. pdf not found | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
377 | 2020 | Semi-supervised Trajectory Understanding with POI Attention for End-to-End Trip Recommendation | Zhou, F., Wu, H., Trajcevski, G., Khokhar, A., Zhang, K. | https://dl.acm.org/doi/10.1145/3378890 | ACM Transactions on Spatial Algorithms and Systems | Journal | 1,3 | 0 | IGNORE. Trip recommendation | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
378 | 2020 | PGR-ELM | A new point-of-interest group recommendation method in location-based social networks | Zhao, X., Zhang, Z., Bi, X., Sun, Y. | https://link.springer.com/article/10.1007/s00521-020-04979-4 | Neural Computing and Applications | Journal | ??? | 1 | 1 | \cite{Zhao2020} | IGNORE. POI group recommendation | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
379 | 2020 | ??? | A MAS-Based Approach for POI Group Recommendation in LBSN | Schiaffino, S., Godoy, D., Pace, J.A.D., Demazeau, Y. | https://link.springer.com/chapter/10.1007/978-3-030-49778-1_19 | International Conference on Practical Applications of Agents and Multi-Agent Systems | Conference | 1 | 1 | ??? | IGNORE. POI group recommendation | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
380 | 2020 | ??? | Travel Route Recommendation via Location-Based Social Network and Skyline Query | Ke, C.-K., Lai, S.-C., Chen, C.-Y., Huang, L.-T. | https://link.springer.com/chapter/10.1007/978-981-15-3250-4_14 | International Conference on Frontier Computing | Conference | ??? | 1 | 0 | IGNORE. Route recommendation | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
381 | 2020 | Using multi-features to partition users for friends recommendation in location based social network | Xin, M., Wu, L. | https://www.sciencedirect.com/science/article/pii/S0306457319303188 | Information Processing & Management | Journal | 1,2 | 6 | IGNORE. Friend recommendation | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
382 | 2020 | GANR | Graph Attentive Network for Region Recommendation with POI- and ROI-Level Attention | Xu, H., Wei, J., Yang, Z., Wang, J. | https://link.springer.com/chapter/10.1007/978-3-030-60259-8_37 | APWeb-WAIM | Conference | 1 | 0 | IGNORE. Region recommendation, not POI | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
383 | 2020 | ---No-Acronym-- | A point of interest recommendation engine with an integrated approach | Jain, P., Mahapatra, A., Mahalakshmi, P. | http://sersc.org/journals/index.php/IJAST/article/view/9357 | International Journal of Advanced Science and Technology | Journal | ??? | 1 | 0 | ??? | Very bad written, no experiments provided. IGNORE | Yes | No | No | No | No | No | No | Yes | No | Yes | No | No | No | No | No | No information | No information | No information | No information | No information | No information | No information | No information | No information | No information | No information | No information | No information | No information | No information | No information | No information | No information | No information | No information | No information | No information | No information | No information | No information | No information | |||||||||||||||||||||||
384 | 2020 | A location-based orientation-aware recommender system using IoT smart devices and Social Networks | Ojagh, S., Malek, M.R., Saeedi, S., Liang, S. | https://www.sciencedirect.com/science/article/abs/pii/S0167739X1930562X | Future Generation Computer Systems | Journal | 1,2 | 4 | IGNORE. Not POI recommendation | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
385 | 2020 | DeepVenue | Deep Learning Driven Venue Recommender for Event-Based Social Networks | Pramanik, S., Haldar, R., Kumar, A., Pathak, S., Mitra, B. | https://ieeexplore.ieee.org/document/8709774 | IEEE Transactions on Knowledge and Data Engineering | Journal | 1 | 1 | \cite{DBLP:journals/tkde/PramanikHKPM20} | More oriented to events. Group recommendation. IGNORE | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
386 | 2020 | Enhancing Multi-factor Friend Recommendation in Location-based Social Networks | Samir, B., El-Tazi, N. | https://ieeexplore.ieee.org/abstract/document/9346530 | International Conference on Data Mining Workshops (ICDMW) | Conference | 1 | 0 | IGNORE. Friend recommendation | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
387 | 2020 | Extracting Travel Demand for Emergency Situations Using Location-Based Social Network Data | Ilil Blum Shem-Tov, Shlomo Bekhor, | https://www.sciencedirect.com/science/article/pii/S2352146520301447 | Transportation Research Procedia | Journal | 2 | 0 | IGNORE. Out of the scope. Emergency situations | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
388 | 2020 | Bi-GTPPP | Exploiting bi-directional global transition patterns and personal preferences for missing POI category identification | Dongbo Xi, Fuzhen Zhuang, Yanchi Liu, Hengshu Zhu, Pengpeng Zhao, Chang Tan, Qing He | https://www.sciencedirect.com/science/article/pii/S089360802030304X | Neural Networks | Journal | 2 | 0 | \cite{DBLP:journals/nn/XiZLZZTH20} | IGNORE. Category prediction | No | No | No | Yes | No | No | Stochastic gradient descent with adam | No | Yes(in the neural network) | No | No | Yes | No | Yes | No | Yes | No | No | No | Yes(80% training, 10% validation, 10% test) | No | No | No | No | Yes(removed users with less than 10 check-ins) | Yes(Tokyo and NewYork) | Ranking | Recall and F1 score | Yes(TOP, TOP2) | No | No | Yes(PRME-G) | Yes(10%) | Check-ins | No | No | Yes | No | No | No | Yes | Prev | No | Foursquare: They refer to Modeling user activity preference by leveraging user spatial temporal characteristics in LBSNs 2015 | |||||||||||||||||||||
389 | 2020 | Exploiting geographical-temporal awareness attention for next point-of-interest recommendation | Liu, T., Liao, J., Wu, Z., Wang, Y., Wang, J. | https://www.sciencedirect.com/science/article/pii/S0925231220300680 | Neurocomputing | Journal | 1,2 | 6 | IGNORE. Same as in \cite{DBLP:conf/mir/LiuLWWW19} | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
390 | 2020 | MCLR- TD | Multi-Context-Aware Location Recommendation Using Tensor Decomposition | Lu, J., Indeche, M.A. | https://ieeexplore.ieee.org/document/9047960 | IEEE Access | Journal | 1 | 1 | \cite{DBLP:journals/access/LuI20} | Yes(CF between items) | Yes(Tensor descomposition) | No | No | No | No | SGD(stochastic gradient descent) | No | No | No | No | Yes(I think yes) | No | No | Yes | Yes | No | No | No | Yes(30% last interaction to test, rest to training and validation) | No | No | No | No | Yes(filtering POIs by category, users with at least 10 different POIs and locations with 5 interactions) | Yes(New York and Tokyo) | Error/Ranking | MAE, RMSE, MRR | No(GeoMF++,ATTF,UZT) | No | No | Yes(GeoMF++) | Yes | Check-ins | No | No | Yes | Yes | No | No | Yes | Post | No | Foursquare and Yelp: for yelp https://www.yelp.com/dataset for Foursquare they refer to Modeling user activity prefer- ence by leveraging user spatial temporal characteristics in LBSNs. Number of POIs do not match other works | ||||||||||||||||||||||
391 | 2020 | ADPM | Attention-Based Dynamic Preference Model for Next Point-of-Interest Recommendation | Zheng, C., Tao, D. | https://link.springer.com/chapter/10.1007/978-3-030-59016-1_63 | International Conference on Wireless Algorithms, Systems, and Applications | Conference | 1 | 0 | \cite{DBLP:conf/wasa/ZhengT20} | No | No | No | Yes | No | No | Adam | No | Yes(embedding of neural networks) | Yes(distance of the POIs) | No | No | No | Yes | Yes | Yes | No | No | No | Yes(last check-in to test, leave one out) | No | No | No | No | Yes(filtering users and POIs with less than10 check-ins) | No | Ranking | Acc and NDCG | Yes(Pop, BPRMF, FPMC, POI2Vec, ST-RNN, SASRec) | Yes(Pop) | Yes(BPRMF) | Yes(POI2Vec) | No | Check-ins | No | Yes | Yes | No | No | Yes(http://spatialkeyword.sce.ntu.edu.sg/eval-vldb17, http://www.yongliu.org/datasets/index.html) | Yes | Post | No | Foursquare: http://spatialkeyword.sce.ntu.edu.sg/eval-vldb17 and Gowalla http://www.yongliu.org/datasets/index.html | ||||||||||||||||||||||
392 | 2020 | ---No-Acronym-- | Location-based social network recommendations with computational intelligence-based similarity computation and user check-in behavior | Elangovan, R., Vairavasundaram, S., Varadarajan, V., Ravi, L. | https://onlinelibrary.wiley.com/doi/full/10.1002/cpe.6106 | Concurrency and Computation: Practice and Experience | Journal | ??? | 1 | 0 | \cite{Elangovan2020} | Yes | No | Yes(power-law, KDE) | No | No | Yes(line 41 of their algorithm) | No | Yes(active + inactive users) | No | Yes | No | No | No | No | Yes | Yes | No | No | Yes(75% for training, rest to test) | No | No | No | No | No | No | No | Ranking | Precision, Recall, F1 | No(MF,SSCI,STU,GTI-BF,GTB) | No | Yes(MF) | Yes(STU) | No | Check-ins | No | No | Yes | No | No | No | Yes | Prev-No filtering | No | Foursquare: no further details | |||||||||||||||||||||
393 | 2020 | ---No-Acronym-- | How to Improve the Recommendation’s Accuracy in POI Domains? | Chaves, L., Silva, N., Carvalho, R., Pereira, A.C.M., Dias, D.R.C., Rocha, L. | https://link.springer.com/chapter/10.1007/978-3-030-58799-4_41 | International Conference on Computational Science and Its Applications | Conference | 1 | 0 | \cite{DBLP:conf/iccsa/ChavesSCPDR20} | Weird, they use reranking approaches or closely related to reranking | No | No | No | No | No | No | Yes(reranking) | No | No | No | No | Yes | No | No | No | No | No | Yes | No | No | Yes(70% of the interactions to training rest to test) | No | No | No | No | No | Yes(POIs with at least 5 visits and users with at least 20 interactions) | Yes(Las Vegas, Phoenix and Charlotte) | Ranking | Precision and Recall | Yes(Pop, UBKNN, USG...) | Yes(Pop) | Yes(UB) | Yes(USG) | No | Check-ins | No | No | No | Yes | No | Yes(https://www.yelp.com/dataset/challenge) | Yes | Post | No | ||||||||||||||||||||
394 | 2020 | TPR-TF | A New Personalized POI Recommendation Based on Time-Aware and Social Influence | Wang, N., Liu, Y., Han, P., Li, X., Li, J. | https://link.springer.com/chapter/10.1007/978-3-030-59635-4_14 | International Conference on Cloud Computing | Conference | 1 | 0 | \cite{DBLP:conf/cloud2/WangLHLL20} | Yes(cosine similarity between frinds) | Yes(Tensor) | No | No | No | No | SGD(stochastic gradient descent) | Yes(Hierarchical clustering) | No | No | Yes(friends) | No | No | No | Yes | Yes | No | No | No | No | Yes(70% training, 10% validation, 20% test) | No | No | No | No | No | Ranking | Precision and Recall | No(RegPMF,UTE+SE, GTAG, USGT, BPR) | No | Yes(BPR) | Yes(UTE+SE) | Yes(10%) | Check-ins | No | Yes | No | No | Brightkite | No | Yes | Prev-No filtering | No | Gowalla and Brightkite, no further details | ||||||||||||||||||||||
395 | 2020 | ??? | POI Recommendation Based on Locality-Specific Seasonality and Long-Term Trends | Stefancova, E., Srba, I. | https://link.springer.com/chapter/10.1007/978-3-030-38919-2_28 | International Conference on Current Trends in Theory and Practice of Informatics | Conference | 1 | 0 | \cite{DBLP:conf/sofsem/StefancovaS20} | Recomend POI reviews, but basically its recommending POIs | No | Yes | No | No | No | No | Warp | No | No | No(They use it in the last part for filtering) | No | Yes | No(they claim to user reviews but they do not exploit them) | No | Yes | Yes | No | No | Yes(last year of reviews) | No | No | No | No | No | Yes(only users with at least 5 reviews and POIs with at least 3 reviews) | Yes | Ranking | Precision and Recall | No | No | No | No | No | Check-ins (reviews) | No | No | No | Yes | No | No | Yes | Prev | No | ||||||||||||||||||||||
396 | 2020 | GLR_GT, GLR_GT_LSTM | GLR: A graph-based latent representation model for successive POI recommendation | Lu, Y.-S., Huang, J.-L. | https://www.sciencedirect.com/science/article/pii/S0167739X19303966 | Future Generation Computer Systems | Journal | 1,2 | 5 | \cite{DBLP:journals/fgcs/LuH20} | successive POI recommendation | No | No | Yes(power-law, GLR_GT, GLR_GT_LSTM) | Yes(GLR_GT_LSTM) | No | Yes(GLR_GT) | SGD(stochastic gradient descent) | No | Yes | Yes(GLR_GT, GLR_GT_LSTM) | No | Yes( GLR_GT, GLR_GT_LSTM) | No | Yes(GLR_GT, GLR_GT_LSTM) | Yes | No | No | Yes(70% training, 10% validation and 10% test) | No | No | No | No | No | No | No | Ranking | Precision and Recall | No | No | No | Yes(POI2Vec) | Yes(10%) | Check-ins | No | Yes | Yes | No | No | No | No | Prev-No filtering | No | Gowalla and Foursquare. No further details | ||||||||||||||||||||||
397 | 2020 | PDPNN | Pdpnn: Modeling user personal dynamic preference for next point-of-interest recommendation | Zhong, J., Ma, C., Zhou, J., Wang, W. | https://link.springer.com/chapter/10.1007/978-3-030-50433-5_4 | International Conference on Computational Science | Conference | 1 | 0 | \cite{DBLP:conf/iccS/ZhongMZW20} | Next-POI recommendation | No | No | No | Yes | No | No | No | No | Yes(generated by neural networks) | Yes | No | No | No | Yes | Yes | Yes | No | No | No information(it is global but not specifically stated) | No | No information(it is global but not specifically stated) | No | No | No | No | Yes(Tokyo, New York, California) | Ranking | Recall | No(Only RNN approaches) | No | No | No | No | Check-ins | No | No | Yes | No | No | No | Yes | Prev-No filtering | No | Foursquare they refer to Participatory cultural mapping based on collective behavior data in location based social networks 2016 and Modeling user activity preference by leveraging user spatial temporal characteristics in LBSNs for TOK and NY | |||||||||||||||||||||
398 | 2020 | MANC | Exploiting multi-attention network with contextual influence for point-of-interest recommendation | Chang, L., Chen, W., Huang, J., Bin, C., Wang, W. | https://link.springer.com/content/pdf/10.1007/s10489-020-01868-0.pdf | Applied Intelligence | Journal | ??? | 1 | 1 | \cite{Chang2020ExploitingMN} | No | Yes(last part, optimized using BPR) | No | Yes | No | No | BPR | No | Yes(neural network) | Yes | Yes(friends) | No | No | No | No | Yes | No | No | Yes(80% training rest to test) | No | No | No | No | No | Yes(users with less than 15 check-ins removed and POIs with less than 10 visited users) | No | Ranking | Precision, Recall, NDCG, MAP | No | No | Yes(BPR) | Yes(LORE, MGMPFM...) | No | Check-ins | No | Yes | No | Yes | No | No | Yes | Prev-No filtering | No | Gowalla and Yelp. No further details | |||||||||||||||||||||
399 | 2020 | GeoSeDMF | Joint Geosequential Preference and Distance Metric Factorization for Point-of-Interest Recommendation | Liu, C., Liu, C., Xin, H., Wang, J., Liu, J., Xu, S. | https://www.hindawi.com/journals/mpe/2020/6582676/ | Mathematical problems in engineering | Journal | ??? | 1 | 0 | \cite{Liu2020} | No | Yes | No | No | No | No | AdaGrad | No | No | Yes | No | No | No | No | No | Yes | No | No | No | Yes(Last POI to test) | No | No | No | No | Yes(remove users and pois with less than 5 check-ins) | Yes(Singapore, New York, California and Nevada) | Ranking | Recall, F1-Score, NDCG | No | No | Yes(BPR) | Yes(GeoMF) | No | POIs | No | Yes | Yes | No | Yes(Foursquare: https:// www.ntu.edu.sg/home/gaocong/datacode.htm, Gowalla: https://www.ntu.edu.sg/home/gaocong/datacode and Instagram: t https://dmis.korea.ac.kr/cape) | Yes | Post | No | Foursquare: https:// www.ntu.edu.sg/home/gaocong/datacode.htm, Gowalla: https://www.ntu.edu.sg/home/gaocong/datacode and Instagram: t https://dmis.korea.ac.kr/cape | ||||||||||||||||||||||
400 | 2020 | STACP | Joint Geographical and Temporal Modeling Based on Matrix Factorization for Point-of-Interest Recommendation | Rahmani, H.A., Aliannejadi, M., Baratchi, M., Crestani, F. | https://link.springer.com/chapter/10.1007/978-3-030-45439-5_14 | European Conference on Information Retrieval | Conference | 1 | 1 | \cite{DBLP:conf/ecir/RahmaniABC20} | No | Yes | No | No | No | Yes(Eq1) | No | No | No | Yes | No | No | No | No | Yes | Yes | No | No | No | Yes(70% training, 10% validation and 20% test) | No | No | No | No | No | Yes(USA) | Ranking | Precision, Recall, NDCG | Yes(Popularity, PMF, PFMMGM...) | Yes(Pop) | No | Yes(Rank-GeoFM, igslr) | Yes(10%) | Check-ins | No | Yes | Yes | No | No | Yes(http://spatialkeyword.sce.ntu.edu.sg/eval-vldb17/) | Yes | Prev-No filtering | Yes(https://github.com/rahmanidashti/STACP) | Yes | No | Yes | Foursquare and Gowalla: They refer to An experimental evaluation of point- of-interest recommendation in location-based social networks 2017 | |||||||||||||||||||
401 | 2020 | GSSM | GSSM: An Integration Model of Heterogeneous Factors for Point-of-Interest Recommendation | Yang, Q., Chen, Y., Luo, P., Zhang, J. | https://link.springer.com/chapter/10.1007/978-981-15-8086-4_2 | International Conference on Artificial Intelligence and Security | Conference | ??? | 1 | 0 | \cite{Yang2020} | Yes | No | Yes(Gaussian model) | No | No | Yes(Eq10) | No | No | No | Yes | Yes | No | No | No | No | Yes | No | No | Yes(80% training, rest to test) | No | No | No | No | No | Yes(users with less than 50 check-ins, removed. Items with less than 50 check-ins also removed) | No | Ranking | Precision and Recall | No(K-means) | No | No | No | No | Check-ins | No | Yes | No | No | No | No | Yes | Prev | No | Gowalla: no further details | |||||||||||||||||||||
402 | 2020 | MTNR | From When to Where: A Multi-task Learning Approach for Next Point-of-Interest Recommendation | Zhong, J., Ma, C., Zhou, J., Wang, W. | https://link.springer.com/chapter/10.1007/978-3-030-59016-1_64 | International Conference on Wireless Algorithms, Systems, and Applications | Conference | 1 | 0 | \cite{DBLP:conf/wasa/ZhongMZW20} | No | No | No | Yes | No | No | SGD(stochastic gradient descent) | No | No | Yes | No | No | No | Yes | Yes | Yes | No | No | No | Yes(90% trajectories to training, rest to test) | No | No | No | No | No | Yes(NY, Tokyo and California) | Ranking | Recall and MAP | No | No | No | Yes(ATST-LSTM) | No | Check-ins | No | No | Yes | No | No | No | Yes | Prev-No filtering | No | Foursquare: from Modeling user activity preference by leveraging user spatial temporal characteristics in LBSNs 2015 | ||||||||||||||||||||||
403 | 2020 | STSAN | Spatio-Temporal Self-Attention Network for Next POI Recommendation | Ni, J., Zhao, P., Xu, J., Fang, J., Li, Z., Xian, X., Cui, Z., Sheng, V.S. | https://link.springer.com/chapter/10.1007/978-3-030-60259-8_30 | Asia-Pacific Web (APWeb) and Web-Age Information Management (WAIM) Joint International Conference on Web and Big Data | Conference | 1 | 0 | \cite{DBLP:conf/apweb/NiZ0FLXCS20} | No | No | No | Yes | No | No | Adam | No | No | Yes | No | No | No | Yes | Yes | Yes | No | No | No | Yes(70% training, 30% test) | No | No | No | No | Yes(users and pois with less than 10 check-ins, removed) | Yes(New York and Tokyo for Foursquare) | Ranking | Recall, NDCG | No | No | No | Yes(STGN) | No | Check-ins | No | Yes | Yes | No | No | Yes(http://snap.stanford.edu/data/loc-gowalla.html for Gowalla and https://sites.google.com/site/yangdingqi/home/foursquare-dataset for Foursquare) | Yes | Post | No | Foursquare: https://sites.google.com/site/yangdingqi/home/foursquare-dataset and Gowalla: http://snap.stanford.edu/data/loc-gowalla.html. 2 | ||||||||||||||||||||||
404 | 2020 | DPR-Geo | DPR-Geo: A POI Recommendation Model Using Deep Neural Network and Geographical Influence | Zeng, J., Tang, H., Wen, J. | https://link.springer.com/chapter/10.1007%2F978-3-030-63836-8_35 | International Conference on Neural Information Processing | Conference | 1 | 0 | \cite{DBLP:conf/iconip/ZengTW20} | No | No | Yes(power-law) | Yes | No | Yes(Eq 18) | Gradient Descent | No | No | Yes | No | No | No | No | No | Yes | No | No | No | No | No | Yes(80% for training, rest to test) | No | No | Yes(users who visited less than 6 POIs removed and POIs with less than 6 users, removed) | Yes(Atlanta and Hawaii) | Ranking | Precision, Recall, F1 | Yes(Pop, MF, BPR) | Yes(Pop) | Yes(BPR, NMF) | No | No | POIs | No | No | Yes | No | No | No | Yes | Prev | No | Foursquare. No further details | ||||||||||||||||||||||
405 | 2020 | SVD++&FMRec | A personalized point-of-interest recommendation system for O2O commerce | Kang, L., Liu, S., Gong, D., Tang, M. | https://link.springer.com/content/pdf/10.1007/s12525-020-00416-5.pdf | Electronic Markets | Journal | ??? | 1 | 2 | \cite{Kang2020} | No | Yes(SVD++) | No | No | No | No | stochastic variance reduced gradient | No | No | Yes | Yes | No | No | No | Yes | Yes | No | No | No | No | No | No | Yes(70% training, 10% validation and 20%test, repeated 5 times) | No | Yes(users and POIs with less than 5 check-ins, removed) | No | Ranking | Precision and Recall | No(MF, PFM,SVD++) | No | Yes(SVD++, FM) | No | Yes(10%) | Check-ins | No | Yes | Yes | No | No | No | Yes | Prev | No | Foursquare: They refer to Location-based and preference-aware recommendation using sparse geo-social network- ing data. 2012, for Gowalla they refer to Friendship and mobility: User movement in location-based social networks 20111 | |||||||||||||||||||||
406 | 2020 | UPEMF | Multi-factor Fusion POI Recommendation Model | Ma, X., Zhu, J., Zhang, S., Zhong, Y. | https://link.springer.com/chapter/10.1007/978-981-15-7984-4_2 | International Conference of Pioneering Computer Scientists, Engineers and Educators | Conference | 1 | 0 | \cite{DBLP:conf/icycsee/MaZZZ20} | No | Yes(Probabilistic MF) | Yes(Probabilistic MF) | No | No | No | Gradient Descent | No | No | Yes | No | No | No | No | No | Yes | No | No | No | No | Yes(80% training 20% test) | No | No | No | Yes(regional blocks) | Yes(USA) | Ranking | Precision and Recall | No | No | Yes(BPR) | Yes(GeoFM) | No | Check-ins | No | Yes | Yes | No | No | No | Yes | Post | No | Foursquare and Gowalla, no further details | ||||||||||||||||||||||
407 | 2020 | SSANet | POI Recommendations Using Self-attention Based on Side Information | Yue, C., Zhu, J., Zhang, S., Ma, X. | https://link.springer.com/chapter/10.1007%2F978-981-15-7984-4_5 | International Conference of Pioneering Computer Scientists, Engineers and Educators | Conference | 1 | 0 | \cite{DBLP:conf/icycsee/YueZZM20} | No | No | Yes(Gaussian kernel for the distance between the POIs) | Yes | No | No | Adam | No | Yes(neural network) | Yes | Yes | No | No | No | Yes | Yes | No | No | No information | No information | No information | No information | No information | No information | No | No | Ranking | Precision, Recall, F1, MAP | No | No | Yes(BPR, WRMF) | Yes(Rank-GeoFM) | No | No information | Yes | Yes | Yes | Yes | No | Yes(https://sites.google.com/site/yangdingqi/home/foursquare-dataset for Foursquare, http://snap.stanford.edu/data/loc-gowalla.html for Gowalla, https://www.yelp.com/dataset/challenge for Yelp | Yes | Post | No | Foursquare, Gowalla and Yelp (https://sites.google.com/site/yangdingqi/home/foursquare-dataset for Foursquare, http://snap.stanford.edu/data/loc-gowalla.html for Gowalla, https://www.yelp.com/dataset/challenge for Yelp | ||||||||||||||||||||||
408 | 2020 | STAR | Modeling POI-Specific Spatial-Temporal Context for Point-of-Interest Recommendation | Wang, H., Shen, H., Cheng, X. | https://link.springer.com/chapter/10.1007/978-3-030-47426-3_11 | Pacific-Asia Conference on Knowledge Discovery and Data Mining | Conference | 1 | 0 | \cite{DBLP:conf/pakdd/WangSC20a} | No | No | No | Yes | No | No | No | No | No | Yes | No | No | No | No | Yes | Yes | No | No | No | Yes(80% training, 10% validation, 10% test) | No | No | No | No | No | No | Ranking | Hit and MRR | No | No | No | Yes(UG, UTG, FPMC) | Yes(10%) | Check-ins | No | Yes | Yes | No | No | No | Yes | Prev-No filtering | No | Foursquare and Gowalla: They refer to An experimental evaluation of point- of-interest recommendation in location-based social networks | ||||||||||||||||||||||
409 | 2020 | JTCR | A Joint Two-Phase Time-Sensitive Regularized Collaborative Ranking Model for Point of Interest Recommendation | Aliannejadi, M., Rafailidis, D., Crestani, F. | https://ieeexplore.ieee.org/document/8661539 | Transactions on Knowledge and Data Engineering | Journal | 1 | 5 | \cite{DBLP:journals/tkde/AliannejadiRC20} | No | No | No | No | No | No | 2 phase, method. Maybe is factorization | Gradient Descent | No | No | Yes | No | No | No | No | Yes | Yes | No | No | No | Yes(70% of the data for each user training, 10% valdiation and 20% test ordered) | No | No | No | No | No | No | Ranking | Precision and NDCG | No(WRMF, GeoMF, IRenMF, Rank-GeoFM, RH-Push, Inf-Push, P-Push) | No | Yes(WRMF) | Yes(IrenMF, GeoMF) | Yes(10% validation) | Check-ins | No | Yes | Yes | No | No | No(They refer to Time-aware point-of-interest recommendation 2013) | Yes | Prev-No filtering | No | Foursquare and Gowalla: They refer to Time-aware point-of-interest recommendation 2013 | |||||||||||||||||||||
410 | 2020 | Rank-FBPR | Points-of-Interest Recommendation Algorithm Based on LBSN in Edge Computing Environment | Cao, K., Guo, J., Meng, G., Liu, H., Liu, Y., Li, G. | https://ieeexplore.ieee.org/document/9031336 | IEEE Access | Journal | 1 | 1 | \cite{DBLP:journals/access/CaoGMLLL20} | No | Yes(BPR-MF) | Yes(BPR-MF) | No | No | Yes(Eq 14) | BPR, MLE(Maximum Likelihood Estimation) | Yes | No | Yes | Yes | No | No | No | No | Yes | No | No | No | No | Yes(80% training, 20% test) | No | No | No | No | No | Ranking | Precision and Recall | No | No | Yes(BPR) | Yes(GeoMF) | No | Check-ins | Yes | No | Yes | Yes | No | No | Yes | Prev-No filtering | No | Foursquare and Yelp: no further details. They claim to use foursquare but the statistics match a dataset from Gowalla | ||||||||||||||||||||||
411 | 2020 | VGMF | VGMF: Visual contents and geographical influence enhanced point-of-interest recommendation in location-based social network | Liu, B., Meng, Q., Zhang, H., Xu, K., Cao, J. | https://onlinelibrary.wiley.com/doi/full/10.1002/ett.3889 | Emerging Telecomunications Tecnologies | Journal | ??? | 1 | 1 | No | Yes | No | Yes(for learning the visual information) | No | No | Gradient Descent | No | No | Yes | No | Yes(visual) | No | No | No | Yes | No | No | No information | No information | No information | No information | No information | No information | No information | Yes(New York) | Ranking | Precision and Recall | No | No | Yes(UB, IB) | Yes(GeoMF) | No information | No information | No information | No | Yes | No | No | Yes(https://github.com/socialsnail/VGMF) | Yes | No | Yes(https://github.com/socialsnail/VGMF) | Yes | No | Yes | Foursquare: No information about its dataset | |||||||||||||||||||
412 | 2020 | BPSL | A Point-of-Interest Recommendation Algorithm Combining Social Influence and Geographic Location Based on Belief Propagation | Li, J., Wang, X., Feng, W. | https://ieeexplore.ieee.org/document/9174725 | IEEE Access | Journal | 1 | 0 | \cite{DBLP:journals/access/LiWF20} | Yes(cosine similarity between friends) | No | Yes(power-law) | No | No | Yes(Eq 21) | No | No | No | Yes | Yes | Yes | No | No | Yes | Yes | No | No | No | No | Yes(80% training, 20% test) | No | No | No | Yes(at least 4 check-ins) | Yes(USA) | Ranking | Precision and Recall | No | No | No | Yes | No | Check-ins | Yes | No | Yes | No | No | No | No | Post | No | Foursquare: They refer to ‘Location recommendation algorithm combining similarity and friend trust in LBSN | ||||||||||||||||||||||
413 | 2020 | ---No-Acronym-- | Point-of-interest (POI) recommender systems for social groups in location based social networks (LBSNs)-Proposition of an improved model | Ngamsa-Ard, S., Razavi, M., Prasad, P.W.C., Elchouemi, A. | http://www.iaeng.org/IJCS/issues_v47/issue_3/IJCS_47_3_01.pdf | International Journal on Computer Science | Journal | ??? | 1 | 0 | ??? | Yes | No | No | No | No | No | No | Yes(k-means) | No | No(They claim to use it but i think its only to filter the last part) | Yes(friends) | No | No | Yes(consecutive time-slots) | Yes | Yes | No | No | No | No | Yes(84% training, 16% test) | No | No | No | No | No | Ranking | Precision | No | No | No | No | No | Check-ins | No | Yes | No | No | Brightkite | No | Yes | Prev-No filtering | No | Gowalla and Brightkite: they refer to CTF-ARA: An adaptive method for POI recommendation based on check-in and temporal features 2017 | |||||||||||||||||||||
414 | 2020 | HI-LDA | Where to go: An effective point-of-interest recommendation framework for heterogeneous social networks | Xiong, X., Qiao, S., Han, N., Xiong, F., Bu, Z., Li, R.-H., Yue, K., Yuan, G. | https://www.sciencedirect.com/science/article/pii/S0925231219313281 | Neurocomputing | Journal | 1,2 | 7 | \cite{DBLP:journals/ijon/XiongQHXBLYY20} | Yes | No | Yes(LDA) | Yes(LDA) | No | No | No | No | Yes | No | Yes | Yes | No(behavior) | Yes | No | No | Yes | No | No | No | No | No | No | Yes(10 fold cross-validation) | No | No | Yes(For Foursquare they use San francisco, but they fuse all datasets into one) | Ranking | Accuracy@k | No | No | No(UPS-CF also takes into account geographical information) | Yes(UOS-CF) | No(cross-validation) | Check-ins | No | No | Yes | No | Twitter, Facebook | No | Yes | Prev-No filtering | No | Foursquare and Twitter and Facebook: no further details. No check-ins stated but reviews | |||||||||||||||||||||
415 | 2020 | SACRA | Adversarial learning to compare: Self-attentive prospective customer recommendation in location based social networks | Li, R., Wu, X., Wang, W. | https://dl.acm.org/doi/abs/10.1145/3336191.3371841 | International Conference on Web Search and Data Mining (WSDM) | Conference | 1 | 4 | \cite{DBLP:conf/wsdm/LiW020} | No | No | Yes(Gaussian mixture model) | Yes | No | No | Yes(ELM, gradient method) | No | No | Yes | No | No | No | No | No | Yes | No | No | No | Yes(Temporal per business, but I would say per user) | No | No | No | No | Yes(For yelp, remove users and customers with less than 20 check-ins, for Foursquare, we remove users and POIs with less than 8) | Yes(Los Angeles and New York for Foursquare, for Yelp, 6 different cities) | Ranking | MAP | No | No | Yes(BPR) | Yes(USG, GeoMF) | Yes(20% for validation) | Check-ins | No | No | Yes | Yes | No | No | Not complete | Prev | No | Foursquare and Yelp: no further details. Not complete statistics | ||||||||||||||||||||||
416 | 2020 | MLRS | Model-Based Location Recommender System Using Geotagged Photos on Instagram | Memarzadeh, M., Kamandi, A. | https://ieeexplore.ieee.org/document/9122274 | International Conference on Web Research | Conference | ??? | 1 | 0 | ??? | No | No | No | Yes(They use skip-gram) | No | No | No | No | Yes(word2Vec) | No | No | No | Yes | No | No | Yes | No | No | No | No | No | No | Yes(weird description but i would classifcy as this) | No | Yes(remove posts with irrelevant hastags) | No | Ranking | Precision, Recall, F1 | No | No | No | No | No(cross-validation) | Check-ins | No | No | Yes | No | No | No | No | No | No | Foursquare: no further details. Statistics not complete | |||||||||||||||||||||
417 | 2020 | GeSSo | Exploiting two-dimensional geographical and synthetic social influences for location recommendation | Liu, J., Zhang, Z., Liu, C., Qiu, A., Zhang, F. | https://www.mdpi.com/2220-9964/9/4/285 | ISPRS International Journal of Geo-Information | Journal | 1 | 3 | \cite{DBLP:journals/ijgi/LiuZLQZ20} | Yes(for users + friends) | No | Yes(geographical component) | No | No | Yes(Eq 15) | No | No | No | Yes | Yes | No | No | No | No | Yes | No | No | No | Yes(for Gowalla and Yelp) | No | Yes(For Foursquare) | No | No | No(but as the datast is extracted from other paper, Im sure it was pre-processed) | No(but according to the maps shown, Foursquare itfocused in a region) | Ranking | Precision, Recall | No | No | No | Yes(USG, Geosoca) | No | Check-ins | No | Yes | Yes | Yes | No | No | Yes | Prev-No filtering | No | Foursquare: They refer to Time-aware point-of-interest recommendation 2013 and Gowalla and Yelp, they refer to An experimental evaluation of point-of-interest recommendation in location-based social networks 2017 | ||||||||||||||||||||||
418 | 2020 | ---No-Acronym-- | Time Distribution Based Diversified Point of Interest Recommendation | Mo, F., Jiao, H., Yamana, H. | https://ieeexplore.ieee.org/document/9095741 | International Conference on Cloud Computing and Big Data Analytics | Conference | ??? | 1 | 0 | ??? | It is not a POI model, but a reranking strategy to improve dicersity | No | No | No | No | No | Yes(Eq 3) | Yes(apart from the hybrid, it is a reranking approach) | No | No | No | No | No | Yes | No | No | Yes | Yes | No | No | No | Yes(70% training, 10% validation, rest to test) | No | No | No | No | Yes(users and pois with less than 10 check-ins, removed) | No | Ranking | Precision, Novelty and Diversity | No | No | No | Yes(USG, LFBCA) | Yes(10%) | Check-ins | No | Yes | No | No | No | Yes(http://snap.stanford.edu/data/loc-gowalla.html) | No | No information | No | Gowalla. They refer to http://snap.stanford.edu/data/loc-gowalla.html | |||||||||||||||||||
419 | 2020 | SEATLE | Few-Shot Learning for New User Recommendation in Location-based Social Networks | Li, R., Wu, X., Wang, W. | https://dl.acm.org/doi/abs/10.1145/3366423.3379994 | The Web Conference | Conference | 1 | 5 | \cite{DBLP:conf/www/LiWW020} | No | No | Yes(Gaussian mixture model) | Yes | No | No | EM(Expectation maximization) | No | No | Yes | No | Yes | No | No | No | Yes | No | No | No information | No information | No information | No information | No information | No information | No | Yes(different cities of the datasets) | Ranking | MAP | No | No | Yes(BPR, WMF) | Yes(USG, GeoMF) | No | No information | No | No | Yes | Yes | No | No | Not complete | No information | No | Foursquare and Yelp: no further details. Not complete statistics | ||||||||||||||||||||||
420 | 2020 | ---No-Acronym-- | POI Recommendation with Interactive Behaviors and User Preference Dynamics Embedding | Yu, Z., Wang, Y., Cao, J., Zhu, G. | https://ieeexplore.ieee.org/document/9137471 | International Conference on Artificial Intelligence and Big Data | Conference | ??? | 1 | 0 | ??? | No | No | Yes | Yes(CNN, graph embeddings) | No | No | No | No | Yes(graph embedding) | No | No | No | No | Yes | Yes(time interval for generating the sequences) | Yes | No | No | No | Yes(80% training for each user, 20% to test) | No | No | No | No | No | No | Ranking | Precision, Recall | No | No | Yes(BPR) | Yes(USG) | No | Check-ins | No | No | Yes | No | Weeplaces | Yes(http://www.yongliu.org/datasets/) | Yes | No | No | Weeplaces:http://www.yongliu.org/datasets/. They refer to Joint representation learning for location-based social networks with multi-grained sequential contexts for FOursquare | |||||||||||||||||||||
421 | 2020 | GAIMC | Geography-Aware Inductive Matrix Completion for Personalized Point-of-Interest Recommendation in Smart Cities | Wang, W., Chen, J., Wang, J., Chen, J., Gong, Z. | https://ieeexplore.ieee.org/document/8887261 | IEEE Internet of Things Journal | Journal | 1 | 13 | \cite{DBLP:journals/iotj/WangCWCG20} | Yes | No | Yes(MF) | Yes(Gaussian model) | No | No | No | Expectation Maximization | No | No | Yes | No | No | No | No | No | Yes | No | No | No | No | Yes | No | No | No | Yes(filter our POIs and users with less than 5 check-ins) | Yes(Singapore, California) | Ranking | AUC | No | No | No | Yes(kMEANS++) | No | Check-ins | No | Yes | Yes | No | No | Yes(https://www.ntu.edu.sg/home/gaocong/datacode.htm) | Yes | Post | No | Foursquare and Gowalla:https://www.ntu.edu.sg/home/gaocong/datacode.htm | |||||||||||||||||||||
422 | 2020 | SPR | Personalized location recommendation by fusing sentimental and spatial context | Zhao, G., Lou, P., Qian, X., Hou, X. | https://www.sciencedirect.com/science/article/pii/S0950705120302161 | Knowledge Based Systems | Journal | 1,2 | 7 | \cite{DBLP:journals/kbs/ZhaoLQH20} | Yes | No | Yes(MF) | No | No | No | No | Gradient Descent | No | No | Yes | No | No | Yes | No | No | Yes | No | No | No | No | No | No | Yes(80% training,20% test, 5 fold cross-validation) | No | No | Yes(6 different cities) | Error | MAE/RMSE | No | No | Yes(BaseMF, BiasMF) | Yes(IRenMF) | No(cross-validation) | Check-ins | No | No | No | No | Yes(https://github.com/rushing-snail/Sentimental-Spatial-POI- Recommendation) | Yes | Prev-No filtering | Yes(https://github.com/rushing-snail/Sentimental-Spatial-POI- Recommendation) | Yes | No | No | ||||||||||||||||||||
423 | 2020 | FGCRec | FGCRec: Fine-Grained Geographical Characteristics Modeling for Point-of-Interest Recommendation | Su, Y., Li, X., Liu, B., Zha, D., Xiang, J., Tang, W., Gao, N. | https://ieeexplore.ieee.org/document/9148797 | International Conference on Communications | Conference | 1 | 0 | \cite{DBLP:conf/icc/Su0LZXTG20} | No | No | Yes | No | No | No | Gradient Ascent | No | No | Yes | No | No | No | No | No | Yes | No | No | No | Yes(70% training, 10% validation, 20% test) | No | No | No | No | Yes(Gowalla, remove users with less than 15 check-ins and POIs with less than 10 check-ins. Foursquare: remove users and POIs with less than 10 check-ins) | No | Ranking | Precision and Recall | Yes | Yes(Pop) | No | Yes(Geosoca, FMFMGM) | Yes(10%) | Check-ins | No | Yes | Yes | No | No | No | Yes | Post | No | Foursquare and Gowalla: They refer to An experimental evaluation of point-of-interest recommendation in location-based social networks 2017 for Gowalla and Lglmf: Local geographical based logis- tic matrix factorization model for poi recommendation forFoursquare) | ||||||||||||||||||||||
424 | 2020 | KEAN | KEAN: Knowledge Embedded and Attention-based Network for POI Recommendation | Zhang, C., Li, T., Gou, Y., Yang, M. | https://ieeexplore.ieee.org/document/9182385 | International Conference on Artificial Intelligence and Computer Applications | Conference | ??? | 1 | 0 | ??? | Yes | No | No | Yes | No | No | No | No | Yes | No | Yes | No | No | Yes | No | Yes | No | No | Yes(80% training, rest to test) | No | No | No | No | No | No | No | Ranking | Precision and Recall | No | No | No | Yes(UCF+G, GeoIE) | No | Check-ins | No | Yes | No | No | No | No | No | Prev-No filtering | No | Gowalla: no further details | |||||||||||||||||||||
425 | 2020 | FGRec | FGRec: A Fine-Grained Point-of-Interest Recommendation Framework by Capturing Intrinsic Influences | Su, Y., Zhang, J.-D., Li, X., Zha, D., Xiang, J., Tang, W., Gao, N. | https://ieeexplore.ieee.org/document/9207571 | International Joint Conference on Neural Networks | Conference | 1 | 0 | \cite{DBLP:conf/ijcnn/SuZ0ZXTG20} | Yes(Eq5) | Yes(Eq 10) | Yes(Eq 10) | No | No | Yes(Eq1) | MLE(Maximum Likelihood Estimation) | No | No | Yes | Yes | Yes | No | No | No | Yes | No | No | No | Yes(70% training, 10% validation, 20% test) | No | No | No | No | Yes(removed users and POIs with less than 10 check-ins) | No | Ranking | Precision, Recall, MAP and NDCG | No | No | No | Yes(Geosoca, FMFMGM) | Yes(10%) | Check-ins | No | No | Yes | Yes | No | No | Yes | Post | No | Foursquare and Yelp: Point-of-interest recommendations: Learning potential check-ins from friends for FOursquare and An experimental evaluation of point-of-interest recommendation in location-based social networks 2017 for Yelp. The stats reported for Yelp matches the one from Foursquare in other source | ||||||||||||||||||||||
426 | 2020 | MMBE | Multi-modal Bayesian embedding for point-of-interest recommendation on location-based cyber-physical–social networks | Huang, L., Ma, Y., Liu, Y., Sangaiah, A.K. | https://www.sciencedirect.com/science/article/pii/S0167739X17310191 | Future Generation Computer Systems | Journal | 1,2 | 12 | \cite{DBLP:journals/fgcs/HuangMLS20} | Yes | No | Yes(latent factor variables, Gaussian LDA) | Yes | Yes(use deep Walk and skip-gram) | No | No | No | No | Yes(gamma distribution for learning embeddings) | Yes | Yes | Yes(topic) | No | Yes | Yes(Not quite sure how do they model them, although they claim to use it) | Yes | No | No | Yes(80% for training and 20% for test all cehckins records) | No | No | No | No | No | Yes(remove users with less than 16 Check-ins and POIs with less than 21 visits) | No | Ranking | Precision and Recall | No(SVDFeature, CoRe, FGLR, TRM, LORE) | No | No | Yes(LORE) | No | Check-ins | No | Yes | No | No | Brightkite | No(They refer to Friendship and mobility: user movement in location-based social networks 2011) | Yes | Post | No | Gowalla and Brighkite: They refer to Friendship and mobility: user movement in location-based social networks 2011 | |||||||||||||||||||||
427 | 2020 | ---No-Acronym-- | An adaptive POI recommendation algorithm by integrating user's temporal and spatial features in LBSNs | Li, M., Zheng, W., Xiao, Y., Jiao, X. | https://dl.acm.org/doi/10.1145/3414274.3414494 | International Conference on Data Science and Information Technology | Conference | 1 | 0 | \cite{DBLP:conf/dsit/LiZXJ20} | Yes | Yes | No | No | No | No | No | No | No | Yes(They use it in the Voronoi diagrams) | No | Yes | No | No | Yes | Yes | No | No | No information | No information | No information | No information | No information | No information | Yes(remove users and POIs with less than 20 check-ins) | No | Ranking | Precision, Recall, NDCG | No | No | No | Yes(PFMMGM) | No | No information | No | Yes | Yes | No | No | No | Not complete | Post | No | Fourquare and Gowalla, no further information. Not complete | ||||||||||||||||||||||
428 | 2020 | GeoSAN | Geography-Aware Sequential Location Recommendation | Lian, D., Wu, Y., Ge, Y., Xie, X., Chen, E. | https://dl.acm.org/doi/10.1145/3394486.3403252 | International Conference on Knowledge Discovery & Data Mining | Conference | 1,3 | 3 | \cite{DBLP:conf/kdd/LianWG0C20} | No | No | No | Yes | No | No | No | No | Yes(in the network) | Yes | No | No | No | Yes | Yes | Yes | No | No | No | Yes(last check-ins to test (from unvisited location), rest to train) | No | No | No | No | Yes(remove users with less than 20 check-ins and remove locations visited less than 10 times) | No | Ranking | Hit and NDCG | No | No | No | Yes(STGN) | No | Check-ins | No | Yes | Yes | No | Brightkite | No | Yes | Post | Yes(https://github.com/libertyeagle/GeoSAN) | Yes | No | No | Foursquare, Gowalla and Brightkite: Friendship and mobility: user movement in location-based social networks 2011 for Gowalla and Brightkite and Revis- iting user mobility and social relationships in lbsns: A hypergraph embedding approach 2019 for Foursquare | |||||||||||||||||||
429 | 2020 | TECF | Trust-Enhanced Collaborative Filtering for Personalized Point of Interests Recommendation | Wang, W., Chen, J., Wang, J., Chen, J., Liu, J., Gong, Z. | https://ieeexplore.ieee.org/document/8930072 | Transactions on Industrial Informatics | Journal | 1 | 14 | \cite{DBLP:journals/tii/WangCWCLG20} | Yes | Yes | No | Yes | Yes(Random Walk, deep ) | Yes(Random Walk, deep ) | Yes | Gradient descent | No | No | Yes | No | No | No | No | Yes | Yes | No | No | No | No | No | Yes(20% of th check-ins for each user to test) | No | No | Yes(remove users and POIs with less than 5 check-ins) | Yes(Singapore for Foursquare and California and Nevada for Gowalla) | Ranking | Precision and Recall | No | No | Yes(BCF) | Yes(GeoSage) | No | POIs | No | Yes | Yes | No | No | Yes(http://www.ntu.edu.sg/home/gaocong/datacode.htm) | Yes | Post | No | Foursquare and Gowalla. They refer to http://www.ntu.edu.sg/home/gaocong/datacode.htm | |||||||||||||||||||||
430 | 2020 | FPR | Cold-start Point-of-interest Recommendation through Crowdsourcing | Mazumdar, P., Patra, B.K., Babu, K.S. | https://dl.acm.org/doi/abs/10.1145/3407182 | ACM Transactions on the Web | Journal | 1 | 1 | \cite{DBLP:journals/tweb/MazumdarPB20} | Yes(Eq8) | No | No | No | No | No(Eq 7 seems but finally not) | No | Yes | No | Yes(Eq 7) | No | No(features extracted from reviews) | Yes | No | No | Yes | No | No | No | No | Yes(10%, 20%...90% training, rest to test) | No | No | No | No | Yes(Pennsylvania) | Ranking | Recall | No | No | No | Yes(USG) | No | Check-ins | Yes | No | No | Yes | No | Yes(https://www.yelp.com/dataset_challenge) | Yes | Prev-No filtering | No | |||||||||||||||||||||||
431 | 2020 | STS | STS: Spatial-Temporal-Semantic personalized location recommendation | Li, W., Liu, X., Yan, C., Ding, G., Sun, Y., Zhang, J. | https://www.mdpi.com/2220-9964/9/9/538 | ISPRS International Journal of Geo-Information | Journal | 1 | 1 | \cite{DBLP:journals/ijgi/LiLYDSZ20} | No | No | Yes | No | No | Yes(Eq 21) | SGD(stochastic gradient descent) | No | No | Yes | No | Yes | No | Yes | Yes | Yes | No | No | No | Yes(70% of the earliest check-ins for training rest to test for each user) | No | No | No | No | No | Yes(Austin, Chicago, Houston, Los Angeles and San Francisco) | Ranking | Precision, MRR | No | No | Yes(BaseMF) | Yes(GeoCF) | No | Check-ins | No | Yes | No | No | No | No | Yes | Prev-No filtering | No | Gowalla: they refer to Personalized Point-of-interest Recommendation by Mining Users’ Preference Transition 2013 | ||||||||||||||||||||||
432 | 2020 | MTPR | MTPR: A multi-task learning based POI recommendation considering temporal check-ins and geographical locations | Xia, B., Bai, Y., Yin, J., Li, Q., Xu, L. | https://www.mdpi.com/2076-3417/10/19/6664 | Applied Sciences | Journal | ??? | 1 | 0 | ??? | No | No | No | Yes | No | No | Gradient descent | No | Yes(The one in the neural network) | Yes | No | No | No | Yes | No(I think they only use temporal information for sequences) | Yes | No | No | Yes(70% training, 15% validation, 15% test) | No | No | No | No | No | Yes(remove POIs and users with less than 10 check-ins) | Yes(Foursquare, Singapore, Gowalla, Nevada) | Ranking | Precision, Recall, F1Score | No | No | No | Yes(GeoIE) | Yes(15%) | Check-ins | No | Yes | Yes | No | No | Yes(https://www.ntu.edu.sg/home/gaocong/datacode.htm | Yes | Post | No | Foursquare and Gowalla: (https://www.ntu.edu.sg/home/gaocong/datacode.htm | |||||||||||||||||||||
433 | 2020 | GGLR | Learning Graph-Based Geographical Latent Representation for Point-of-Interest Recommendation | Chang, B., Jang, G., Kim, S., Kang, J. | https://dl.acm.org/doi/10.1145/3340531.3411905 | International Conference on Information & Knowledge Management | Conference | 1 | 2 | \cite{DBLP:conf/cikm/ChangJKK20} | No | No | No | Yes | No | No | No | No | No | Yes | No | No | No | No | No | Yes | No | No | Yes(70% training, 20% validation and 10% test) | No | No | No | No | No | Yes(remove users and POIs with less than 10 check-ins in all datasets) | No | Ranking | Precision, Recall, MAP, NDCG | No | No | No | Yes(IrenMF, GeoMF) | Yes(20%) | Check-ins | No | Yes | Yes | Yes | No | Yes(http://snap.stanford.edu/data/loc-gowalla.htm for Gowalla and https:// www.yelp.com/dataset challenge for Yelp) | Yes | Post | No | Gowalla: http://snap.stanford.edu/data/loc-gowalla.htm. Yelp: https:// www.yelp.com/dataset challenge for Yelp. No info for FOursquare | ||||||||||||||||||||||
434 | 2020 | STP-UDGAT | STP-UDGAT: Spatial-Temporal-Preference User Dimensional Graph Attention Network for Next POI Recommendation | Lim, N., Hooi, B., Ng, S.-K., Wang, X., Goh, Y.L., Weng, R., Varadarajan, J. | https://dl.acm.org/doi/10.1145/3340531.3411876 | International Conference on Information & Knowledge Management | Conference | 1,3 | 2 | \cite{DBLP:conf/cikm/LimHNWGWV20} | No | No | No | Yes | Yes(Random Walk, deep ) | No | No | No | Yes(in the neural network) | Yes | No | No | No | Yes | Yes | Yes | No | No | No | Yes(70% for each user to training, rest to test) | No | No | No | No | Yes(remove users with less than 10 visits) | No | Ranking | Accuracy, Map | Yes | Yes(TOP) | Yes(MF) | No | No | Check-ins | Yes | Yes | Yes | No | Brightkite and others | Yes(http://snap.stanford.edu/data/loc-gowalla.html, http://snap.stanford.edu/data/loc-brightkite.html, https://sites.google.com/site/yangdingqi/home, https://www.start.umd.edu/gtd, https://www.ntu.edu.sg/home/gaocong/datacode.htm) | Yes | Post | No | Gowalla: http://snap.stanford.edu/data/loc-gowalla.html. Brighkite: http://snap.stanford.edu/data/loc-brightkite.html, Foursquare: https://sites.google.com/site/yangdingqi/home, GTD: https://www.start.umd.edu/gtd Another Foursquare: https://www.ntu.edu.sg/home/gaocong/datacode.htm) | ||||||||||||||||||||||
435 | 2020 | HGMAP | Hybrid graph convolutional networks with multi-head attention for location recommendation | Zhong, T., Zhang, S., Zhou, F., Zhang, K., Trajcevski, G., Wu, J. | https://link.springer.com/article/10.1007/s11280-020-00824-9 | World Wide Web | Journal | 1 | 1 | \cite{DBLP:journals/www/ZhongZZZTW20} | No | No | No | Yes | No | No | No | No | No | Yes | Yes | No | No | No | No | Yes | No | No | No | No | No | Yes(80% for training, rest to test) | No | No | Yes(remove users and POIs with less than 20 check-ins, for Foursquare, remove users and POIs with less than 10 check-ins) | No | Ranking | Precision, Recall, MAP | No | No | Yes(BPRMF) | Yes(MGMMF, IRenMF) | No | Check-ins | No | Yes | Yes | Yes | No | No(only for Yelp: https://www.yelp.com/dataset/challenge) | Yes | Post | No | Yelp: https://www.yelp.com/dataset/challenge. Rest no information | ||||||||||||||||||||||
436 | 2020 | HAM-POIRec | Efficient point-of-interest recommendation with hierarchical attention mechanism | Pang, G., Wang, X., Hao, F., Wang, L., Wang, X. | https://www.sciencedirect.com/science/article/pii/S1568494620304750 | Applied Soft Computing | Journal | 1,2 | 2 | \cite{DBLP:journals/asc/PangWHWW20} | No | No | No | Yes | No | No | No | No | No | No | No | No | Yes | Yes | Yes | Yes | No | No | No | No | Yes(80% training, 10% validation,10%test) | No | No | No | No | No | Ranking | MRR, IoU, MAP | No | No | No | Yes(MGMPMF, LRT) | Yes(10%) | Check-ins | Yes | No | No | Yes | No | Yes(https://www.yelp.com/dataset) | Yes | Prev-No filtering | No | |||||||||||||||||||||||
437 | 2020 | ---No-Acronym-- | Exploring Geographic Information Effects for POI Recommendation in LBSNs | Liu, S., Zheng, W., Xiao, Y. | https://iopscience.iop.org/article/10.1088/1742-6596/1651/1/012117 | Journal of Physics: Conference Series | Journal | ??? | 1 | 0 | \cite{Liu_2020} | Yes | No | No | No | No | Yes | No | No | No | Yes | No | Yes | No | No | No | Yes | No | No | No | No | Yes(80% training, rest to test) | No | No | No | No | Yes(Tokyo and NewYork) | Ranking | Precision, Recall, NDCG | No | No | No | Yes(PMFMGM) | No | Check-ins | No | No | Yes | No | No | No | Not complete | Prev-No filtering | No | Foursquare: no further details | |||||||||||||||||||||
438 | 2020 | ---No-Acronym-- | Providing privacy preserving in next POI recommendation for Mobile edge computing | Kuang, L., Tu, S., Zhang, Y., Yang, X. | Journal of Cloud Computing | Journal | 1 | 2 | \cite{DBLP:journals/jcloudc/KuangTZY20} | No | No | Yes(HMM) | No | No | No | Expectation Maximization | No | No | Yes | No | No | No | Yes | No | Yes | No | No | No | Yes(10 months to test, rest to training) | No | No | No | No | No | No | Ranking | Precision | No | No | No | Yes(PRME) | No | Check-ins | No | Yes | Yes | No | No | No | Yes | Prev-No filtering | No | Foursquare: They refer to Spatial-aware hierarchical collaborative deep learning for POI recommendation 2017. Gowalla: no further details | |||||||||||||||||||||||
439 | 2020 | UP2VEC | Heterogeneous graph-based joint representation learning for users and POIs in location-based social network | Yaqiong Qiao, Xiangyang Luo, Chenliang Li, Hechan Tian, Jiangtao Ma | https://www.sciencedirect.com/science/article/pii/S0306457319305114 | Information Processing & Management | Journal | 2 | 0 | \cite{DBLP:journals/ipm/QiaoLLTM20} | No | No | Yes(They say they use power-laws) | Yes | Yes(Random Walk, deep ) | No | Gradient descent | No | Yes(in the neural network) | Yes | Yes(friends) | No | No | Yes | Yes | Yes | No | No | No | Yes(70% training, 10% validation, 10% test). IT DOES NOT SUM 100 | No | No | No | No | No | Yes | Ranking | Accuracy | No | No | No | Yes(Rank-GeoFM) | Yes(10%) | Check-ins | Yes | Yes | Yes | No | No | No | Yes | Prev-No filtering | No | Gowalla: They refer to A random walk around the city: New venue recommendation in location-based social networks 2012. Foursquare: They refer to LCARS: A location-content-aware recommender system 2013 | ||||||||||||||||||||||
440 | 2020 | Dystal | Dynamic discovery of favorite locations in spatio-temporal social networks | Xi Xiong, Fei Xiong, Jun Zhao, Shaojie Qiao, Yuanyuan Li, Ying Zhao, | https://www.sciencedirect.com/science/article/pii/S0306457320308323 | Information Processing & Management | Journal | 2 | 0 | \cite{DBLP:journals/ipm/XiongXZQLZ20} | No | No | Yes | Yes | No | No | asynchronous stochastic gradient algorithm (ASGD) | No | Yes(in the neural network) | Yes | Yes | No | Yes | No | Yes | Yes | No | No | No | No | No | No | Yes(10 fold cross-validation) | No | Yes(remove users and POIs wih less than 10 reviews for Yelp) | Yes(Singapore for Foursquare and LasVegas for Yelp) | Ranking | Precision, Recall and F1 | No | No | No | Yes(TGSC-PMF) | No(cross-validation) | Check-ins | No | No | Yes | Yes | No | No(only for Yelp: https://www.yelp.com/dataset/challenge) | Not complete | Prev | No | Foursquare: no further information. Yelp: https://www.yelp.com/dataset/challenge. Not complete statistics | ||||||||||||||||||||||
441 | 2020 | CRCF | A Contextual Recurrent Collaborative Filtering framework for modelling sequences of venue check-ins | Jarana Manotumruksa, Craig Macdonald, Iadh Ounis | https://www.sciencedirect.com/science/article/pii/S0306457319301876 | Information Processing and Management | Journal | 2 | 0 | \cite{Manotumruksa2019} | No | No | No | Yes | No | No | No | No | Yes(but in the DNN) | Yes | No | No | No | Yes | Yes | Yes | No | No | No | Yes(leave one out methodology) | No | No | No | No | Yes(removed venues with less than 10 Check-ins) | No | Ranking | Hit Rate and NDCG, Risk and Reward | Yes(MostPop. MostVisit, RecentVisit, MF, BPR, GeoBPR, RNN, DREAM, NeuMF, DRCF, STELLAR, CARA) | Yes(Pop) | Yes(BPR) | Yes(GeoBPR, STELLAR) | No | POIs(they create the test with other 100 venues that the user has not visited) | Yes(users with less than 10 Check-ins) | No | Yes | Yes | Brightkite | Yes(https://archive.org/details/201309_foursquare_dataset_umn for Foursquare, https://snap.stanford.edu/data/ for Brighkite, for Yelp https://www.yelp.com/dataset/challenge) | Yes | Post | No | Brighkite, Foursquare, Yelp: https://snap.stanford.edu/data/ for Brightkite, https://archive.org/details/201309_foursquare_dataset_umn for Foursquare and https://www.yelp.com/dataset_challenge for Yelp | ||||||||||||||||||||||
442 | 2020 | MTAS | A point-of-interest suggestion algorithm in Multi-source geo-social networks | Xi Xiong, Shaojie Qiao, Yuanyuan Li, Nan Han, Guan Yuan, Yongqing Zhang, | https://www.sciencedirect.com/science/article/pii/S0952197619302994 | Engineering Applications of Artificial Intelligence | Journal | 2 | 0 | \cite{DBLP:journals/eaai/XiongQLHYZ20} | No | Yes(latent probabilistic approach) | Yes(latent probabilistic approach) | No | No | No | No | No | No | Yes | Yes | No | Yes | No | Yes | Yes | No | No | No | No | No | No | Yes(10 fold cross-validation) | No | Yes(removed users without anchor links) | Yes(San Francisco) | Ranking | Accuracy | No | No | No | Yes(UPS-CF) | No(cross-validation) | Check-ins | No | No | Yes | No | Yes(Twitter and Facebook) | No | Yes | Post | No | No information about its dataset | ||||||||||||||||||||||
443 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
444 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
445 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
446 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
447 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
448 | Temporal (sum this) | Random (sum this) | Other | Using ranking metrics | Using Check-in Split | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
449 | Sim CF (using a structure similar to KNN) | Factorization | Probabilistic | Deep Learning | Social Graph/Link-analysis | Hybrid | Other | Geographical | Social | User/Item content information | Textual | Sequential | Temporal | Offline | Online | User Study | Temporal: Global-Temporal split with a global moment of split | Temporal: User-Temporal split for every user | Random Global: random split in a global manner | Random User: random split for every user | n-foldGlobal | n-foldUser | Other(e.g. different cities for training/test, user case study etc) | Filter users/items | Independent cities/Filtered regions. SPECIFICALLY STATED IN THE PAPER. If they refer those details to other paper it does not count, it needs to be specifically stated | Error/Ranking Metrics | Metrics | Test against classic Rec algorithms (Yes if Pop Is being tested as baseline) | Classic Non personalized baseline (Pop or random) | Classic personalized recommender (UB, IB, BPR, MFs, but not PMF. PMF is not classical. Classical MFs are SVDs, Non matrix factorization, o weighted matrix factorization) | Geographical baseline | Use validation subset | Split by Check-ins or locations/POIS | Cold start analysis? | ||||||||||||||||||||||||||||||||||||||||||
450 | Color of paper | Description | MOST REPRESENTATIVE PAPERS (SUM) | 9 | 27 | 26 | 8 | 6 | 10 | 1 | 38 | 14 | 11 | 5 | 13 | 18 | 45 | 0 | 0 | 8 | 12 | 7 | 12 | 2 | 0 | 1 | 27 | 28 | 43 | 3 | 32 | 33 | 9 | 27 | 7 | |||||||||||||||||||||||||||||||||||||||||
451 | This paper is not recommending POIs (task) or it is not on the scope. It is ignored | Papers per year. Conference vs Journal | ALL PAPERS (SUM) | 90 | 141 | 139 | 66 | 44 | 98 | 19 | 218 | 116 | 108 | 42 | 73 | 134 | 306 | 0 | 5 | 51 | 67 | 62 | 64 | 22 | 4 | 14 | 171 | 144 | 294 | 35 | 173 | 198 | 53 | 205 | 38 | Gowalla | Foursquare | Yelp | Brightkite | Other | ||||||||||||||||||||||||||||||||||||
452 | I could not obtain the pdf of the paper. Ignored paper | Most Representatives | 30 | 34 | 4 | 6 | 8 | 14 | 8 | 5 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
453 | Repeated (for example same title in workshop and a journal with the same authors and same or very similar algorithm proposal, methodology and computations). | Conference papers (the ones in wich column g is conference and are not ignored) | Journal papers (the ones in wich column g is journal and are not ignored) | Temporal | Random | Using Error metrics | Using POI split | Total Number of papers | 156 | 199 | 54 | 40 | 43 | 81 | 41 | 14 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
454 | Ignored paper due to repetition | 2011 | 1 | 0 | 20 | 21 | 5 | 16 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
455 | 2012 | 4 | 0 | 118 | 152 | 25 | 71 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
456 | 2013 | 16 | 0 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
457 | 2014 | 12 | 2 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
458 | 2015 | 22 | 10 | Global split (sum of global temporal, global random and global nfold) | Fix split(sum of fix temporal, fix random and fix nfold) | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
459 | 2016 | 39 | 10 | 17 | 24 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
460 | 2017 | 26 | 13 | 135 | 135 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
461 | 2018 | 26 | 16 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
462 | 2019 | 30 | 29 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
463 | 2020 | 26 | 28 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
464 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
465 | Total Conference | Total Journals | Total Processed Papers (conferences + journals) | Papers per Year. Algorithm methodology | Sim CF (using a structure similar to KNN) | Factorization | Probabilistic | Deep Learning | Social Graph/Link-analysis | Hybrid | Other | Papers per Year. Algorithm methodology | Geographical | Social | User/Item content information | Textual | Sequential | Temporal | Temporal: Global-Temporal split with a global moment of split | Temporal: User-Temporal split for every user | Random Global: random split in a global manner | Random User: random split for every user | n-foldGlobal | n-foldUser | Other(e.g. different cities for training/test, user case study etc) | Papers per Year. Algorithm methodology | Filter users/items | Independent cities/Filtered regions. SPECIFICALLY STATED IN THE PAPER. If they refer those details to other paper it does not count, it needs to be specifically stated | Error/Ranking Metrics -> Ranking | Metrics | Test against classic Rec algorithms (Yes if Pop Is being tested as baseline) | Classic Non personalized baseline (Pop or random) | Classic personalized recommender (UB, IB, BPR, MFs, but not PMF. PMF is not classical. Classical MFs are SVDs, Non matrix factorization, o weighted matrix factorization) | Geographical baseline | Use validation subset | Split by Check-ins or locations/POIS | Cold start analysis? | |||||||||||||||||||||||||||||||||||||||
466 | 202 | 108 | 310 | 2011 | 1 | 0 | 1 | 0 | 0 | 1 | 0 | 2011 | 1 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 2011 | 0 | 0 | 1 | 0 | 0 | 0 | 1 | 1 | 0 | 0 | 1 | |||||||||||||||||||||||||||||||||||||||
467 | 2012 | 3 | 0 | 1 | 0 | 2 | 1 | 0 | 2012 | 3 | 2 | 2 | 0 | 0 | 0 | 1 | 0 | 1 | 0 | 0 | 0 | 1 | 2012 | 1 | 3 | 3 | 0 | 1 | 1 | 4 | 3 | 0 | 0 | 0 | ||||||||||||||||||||||||||||||||||||||||||
468 | 2013 | 6 | 8 | 10 | 0 | 2 | 5 | 1 | 2013 | 8 | 7 | 4 | 3 | 2 | 5 | 4 | 1 | 4 | 5 | 0 | 1 | 1 | 2013 | 8 | 8 | 14 | 0 | 1 | 1 | 10 | 8 | 1 | 0 | 3 | ||||||||||||||||||||||||||||||||||||||||||
469 | 2014 | 5 | 5 | 4 | 1 | 3 | 3 | 0 | 2014 | 10 | 7 | 5 | 1 | 1 | 5 | 3 | 0 | 1 | 7 | 1 | 0 | 1 | 2014 | 5 | 8 | 13 | 0 | 2 | 2 | 9 | 7 | 2 | 0 | 0 | ||||||||||||||||||||||||||||||||||||||||||
470 | 2015 | 9 | 20 | 17 | 1 | 4 | 12 | 2 | 2015 | 22 | 15 | 10 | 8 | 2 | 13 | 9 | 4 | 3 | 9 | 2 | 0 | 2 | 2015 | 16 | 15 | 31 | 0 | 3 | 5 | 23 | 18 | 5 | 0 | 5 | ||||||||||||||||||||||||||||||||||||||||||
471 | 2016 | 18 | 24 | 22 | 1 | 7 | 16 | 3 | 2016 | 29 | 14 | 23 | 7 | 8 | 25 | 13 | 7 | 9 | 6 | 3 | 1 | 7 | 2016 | 29 | 23 | 43 | 0 | 3 | 3 | 25 | 23 | 3 | 0 | 5 | ||||||||||||||||||||||||||||||||||||||||||
472 | 2017 | 12 | 19 | 14 | 5 | 10 | 12 | 2 | 2017 | 24 | 12 | 19 | 6 | 11 | 12 | 2 | 7 | 10 | 12 | 3 | 1 | 0 | 2017 | 23 | 19 | 37 | 0 | 2 | 4 | 23 | 24 | 4 | 0 | 4 | ||||||||||||||||||||||||||||||||||||||||||
473 | 2018 | 10 | 23 | 19 | 9 | 4 | 13 | 4 | 2018 | 32 | 19 | 13 | 5 | 8 | 19 | 5 | 7 | 12 | 9 | 4 | 1 | 2 | 2018 | 21 | 15 | 42 | 0 | 5 | 5 | 24 | 31 | 10 | 0 | 3 | ||||||||||||||||||||||||||||||||||||||||||
474 | 2019 | 13 | 25 | 27 | 22 | 8 | 22 | 5 | 2019 | 44 | 22 | 20 | 5 | 24 | 26 | 4 | 21 | 12 | 11 | 3 | 0 | 0 | 2019 | 39 | 27 | 58 | 0 | 4 | 7 | 32 | 38 | 13 | 0 | 8 | ||||||||||||||||||||||||||||||||||||||||||
475 | 2020 | 13 | 17 | 24 | 27 | 4 | 13 | 2 | 2020 | 45 | 17 | 12 | 7 | 17 | 29 | 10 | 20 | 10 | 4 | 6 | 0 | 0 | 2020 | 29 | 26 | 52 | 0 | 7 | 7 | 22 | 45 | 15 | 0 | 9 | ||||||||||||||||||||||||||||||||||||||||||
476 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
477 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
478 | Years | 2011 | 2012 | 2013 | 2014 | 2015 | 2016 | 2017 | 2018 | 2019 | 2020 | Years | 2011 | 2012 | 2013 | 2014 | 2015 | 2016 | 2017 | 2018 | 2019 | 2020 | Papers per Year. Algorithm methodology | Temporal(both CC and FIx) | Random(both nfold and classic random CC or Fix) | Other | ||||||||||||||||||||||||||||||||||||||||||||||||||
479 | SimCF | 1 | 3 | 6 | 5 | 9 | 18 | 12 | 10 | 13 | 13 | Geographical | 1 | 3 | 8 | 10 | 22 | 29 | 24 | 32 | 44 | 45 | 2011 | 0 | 1 | 0 | ||||||||||||||||||||||||||||||||||||||||||||||||||
480 | Factorization | 0 | 0 | 8 | 5 | 20 | 24 | 19 | 23 | 25 | 17 | Social | 1 | 2 | 7 | 7 | 15 | 14 | 12 | 19 | 22 | 17 | 2012 | 1 | 1 | 1 | ||||||||||||||||||||||||||||||||||||||||||||||||||
481 | Probabilistic | 1 | 1 | 10 | 4 | 17 | 22 | 14 | 19 | 27 | 24 | Content | 0 | 2 | 4 | 5 | 10 | 23 | 19 | 13 | 20 | 12 | 2013 | 5 | 10 | 1 | ||||||||||||||||||||||||||||||||||||||||||||||||||
482 | DeepLearning | 0 | 0 | 0 | 1 | 1 | 1 | 5 | 9 | 22 | 27 | Textual | 0 | 0 | 3 | 1 | 8 | 7 | 6 | 5 | 5 | 7 | 2014 | 3 | 9 | 1 | ||||||||||||||||||||||||||||||||||||||||||||||||||
483 | SocialGraph | 0 | 2 | 2 | 3 | 4 | 7 | 10 | 4 | 8 | 4 | Sequential | 0 | 0 | 2 | 1 | 2 | 8 | 11 | 8 | 24 | 17 | 2015 | 13 | 14 | 2 | ||||||||||||||||||||||||||||||||||||||||||||||||||
484 | Hybrid | 1 | 1 | 5 | 3 | 12 | 16 | 12 | 13 | 22 | 13 | Temporal | 0 | 0 | 5 | 5 | 13 | 25 | 12 | 19 | 26 | 29 | 2016 | 20 | 19 | 7 | ||||||||||||||||||||||||||||||||||||||||||||||||||
485 | Other | 0 | 0 | 1 | 0 | 2 | 3 | 2 | 4 | 5 | 2 | 2017 | 9 | 26 | 0 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
486 | 2018 | 12 | 26 | 2 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
487 | 2019 | 25 | 26 | 0 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
488 | 2020 | 30 | 20 | 0 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
489 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
490 | Papers per Year. Algorithm methodology | 2011 | 2012 | 2013 | 2014 | 2015 | 2016 | 2017 | 2018 | 2019 | 2020 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
491 | Temporal | 0 | 1 | 5 | 3 | 13 | 20 | 9 | 12 | 25 | 30 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
492 | Random | 1 | 1 | 10 | 9 | 14 | 19 | 26 | 26 | 26 | 20 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
493 | Other | 0 | 1 | 1 | 1 | 2 | 7 | 0 | 2 | 0 | 0 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
494 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
495 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
496 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
497 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
498 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
499 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
500 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
501 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
502 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
503 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
504 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
505 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
506 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
507 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
508 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
509 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
510 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
511 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
512 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
513 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
514 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
515 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
516 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
517 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
518 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||