Point-of-Interest Recommendation using Supervised Link Prediction

Abstract

Location-based social networks are a platform that users use to share their information. One of the reasons for using this platform is that POIs are checked in by users and in a sense that POI is recommended to next visitors. The aim of this paper is to find out whether people from the same country visit the same POIs and recommend them accordingly. For this purpose, firstly, a bi-partite network consisting of users and POIs is created. Then, new POIs are recommended to the user regarding the link prediction method over this network. The experimental results on real data obtained from the Foursquare social network showed that the proposed method achieved high accuracy values. © 2021 IEEE.