Point-of-Interest (POI) recommendation has been an important topic on Location-Based Social Networks (LBSN). It could recommend the POI point for users that they have never been. During the latest research, when adding geographical influence, recent research always pick up all the POI points to learn the influence it makes to the users. However, this may reduce the precision of experiment, for it does not take into consideration the reason that influences users in their frequent check-in activity region. To solve this problem, we propose a new POI recommending approach with the activity region, named Activity region Bayesian Personalized Ranking (ABPR), which adds geographical influence into the basket of BPR. This paper outlines the experiments done with Gowalla and Foursquare datasets to demonstrate the effectiveness and advantage of our approach.
CITATION STYLE
Gao, J., & Yang, Y. (2018). ABPR-- A new way of point-of-interest recommendation via geographical and category influence. In Communications in Computer and Information Science (Vol. 902, pp. 96–107). Springer Verlag. https://doi.org/10.1007/978-981-13-2206-8_9
Mendeley helps you to discover research relevant for your work.