Point-of-Interest Recommendation in Location-Based Social Networks

Abstract

This article investigates the problem of proposing new locations to users who have previously been informed about location-based social networks. In the study, user profiles (past location information), the popularity of the place and the time of preference of the place were obtained. Using this information, a different place is offered to the users. However, when users have a different location (a different city or country), it is provided to recommend one of the point of interests designated by categories according to the pre-made check-in details. Social media has been a component of daily life of users. Applications such as Foursquare/Swarm constitute the data source of the communication that geographically contains location data. This study analyzes the personal behaviors of the users using social networks for the spaces and reveals the characteristics of the spaces and people by past notification activities through the applications made in the past. The experiments on real foursquare data showed the usability of the proposed method and achieved results with high accuracy. © 2019 IEEE.