Next point-of-interest recommendation via a category-aware Listwise Bayesian Personalized Ranking

In recent years, location-based social networks (LBSNs) have attracted much attention, and next point-of-interest (POI) recommendation has become an important task for LBSNs. However, previous efforts suffer from the high computational complexity, besides the transition pattern between POIs has not been well studied. In this paper, we proposed a two-fold approach for next POI recommendation. First, we predict the next category by using a third-rank tensor optimized by a Listwise Bayesian Personalized Ranking (LBPR) approach. Specifically, we introduce two functions, namely Plackett–Luce model and cross entropy, to generate the likelihood of a ranking list for posterior computation. The optimization criteria of LBPR are derived from a perspective of Bayesian analysis for ranking problems, and we formulate the listwise loss functions by two probability model. Then POI candidates filtered by the predicated category are ranked based on the spatial influence and category ranking influence. By explicit usage of category information to infer the user transition pattern in category-level, the proposed models are ideally better with the next new POI recommendation problem. The experiments on two real-world datasets demonstrate the significant improvements of our methods over several state-of-the-art methods.