POI Recommendation Based on Locality-Specific Seasonality and Long-Term Trends
This work deals with time-aware recommender systems in a domain of location-based social networks, such as Yelp or Foursquare. We propose a novel method to recommend Points of Interest (POIs) which considers their yearly seasonality and long-term trends. In contrast to the existing methods, we model these temporal aspects specifically for individual geographical localities instead of globally. According to the results achieved by the experimental evaluation on Yelp dataset, locality-specific seasonality can significantly improve the recommendation performance in comparison to its global alternative. We found out that it is helpful mostly within recommendations for highly-active users (it has a smaller influence for the novice users) and as expected, in localities with a strong seasonal weather variation. Another interesting finding is that in contrast to seasonality, we did not observe an improvement in case of locality-specific long-term trends.
KeywordsRecommender systems Points of interest Time-aware recommendation Seasonality Long-term trends
This work was partially supported by the Slovak Research and Development Agency under the contracts No. APVV-15-0508 and APVV SK-IL-RD-18-0004, by the Scientific Grant Agency of the Slovak Republic under the contracts No. VG 1/0667/18 and VG 1/0725/19, and by the student grant provided by Softec Pro Society.
- 3.Baltrunas, L., Ludwig, B., Ricci, F.: Matrix factorization techniques for context aware recommendation. In: Proceedings of the Fifth ACM Conference on Recommender Systems, pp. 301–304. RecSys 2011. ACM, New York (2011). https://doi.org/10.1145/2043932.2043988
- 4.Baltrunas, L., Ricci, F.: Context-dependent recommendations with items splitting, vol. 560, pp. 71–75 (2010)Google Scholar
- 5.Bao, J., Zheng, Y., Mokbel, M.F.: Location-based and preference-aware recommendation using sparse geo-social networking data. In: Proceedings of the 20th International Conference on Advances in Geographic Information Systems, pp. 199–208. SIGSPATIAL 2012, ACM, New York (2012). https://doi.org/10.1145/2424321.2424348
- 6.Basilico, J., Raimond, Y.: DéJà Vu.: The Importance of Time and Causality in Recommender Systems. In: Proceedings of the Eleventh ACM Conference on Recommender Systems, p. 342 (2017). https://doi.org/10.1145/3109859.3109922
- 7.Kula, M.: Metadata embeddings for user and item cold-start recommendations. CoRR abs/1507.08439 (2015)Google Scholar
- 8.Peña, F.J.: Unsupervised context-driven recommendations based on user reviews. In: Proceedings of the Eleventh ACM Conference on Recommender Systems - RecSys 2017, pp. 426–430 (2017). https://doi.org/10.1145/3109859.3109865
- 11.Zhang, J.D., Chow, C.Y.: GeoSoCa: exploiting geographical, social and categorical correlations for point-of-interest recommendations. In: Proceedings of the 38th International ACM SIGIR Conference on Research and Development in Information Retrieval. SIGIR 2015, pp. 443–452. ACM, New York (2015). https://doi.org/10.1145/2766462.2767711
- 12.Zhao, S., Zhao, T., King, I., Lyu, M.R.: Geo-teaser: geo-temporal sequential embedding rank for point-of-interest recommendation. In: WWW (2017)Google Scholar