Behavior-based location recommendation on location-based social networks

Abstract

Location recommendation methods on location-based social networks (LBSN) discover the locational preference of users along with their spatial movement patterns from users’ check-ins and provide users with recommendations of unvisited places. The growing popularity of LBSNs and abundance of shared location information has made location recommendation an active research area in the recent years. However, the existing methods suffer from one or more deficiencies such as data sparsity, cold-start users, ignoring users’ specific spatial and temporal behaviors, not utilizing the shared behaviors of the users. In this paper, we propose a novel location recommendation method, namely Behavior-based Location Recommendation (BLR). BLR recommends a location to a user based on the users’ repetitive behaviors and behaviors of similar users. Additionally, to better integrate the spatial information, BLR has two spatial components, a user-based spatial component to find the spatial preferences of the user, and a behavior-based spatial\xa0component to find locations of interest for different behaviors. Experimental studies on three real-world datasets show that BLR produces better location recommendations and can effectively address data sparsity and cold-start problems.