Evaluation of Social, Geography, Location Effects for Point-of-Interest Recommendation

摘要

Recently, location-based social network service has become very popular. Point of interests (POI) recommendation service is a promising and interesting research problem. In POI recommendation, numbers of locations are more than numbers of users, so it is a challenge to recommend interest locations from amount of location sets. Our idea is to personally incorporate user preference, social influence and attraction of locations in the recommendation. First, we use geographic influence for candidate selection. Furthermore, we propose a unified POI recommendation framework CLW, which fuses user preference to a POI with social influence and attraction of locations. In addition, we discuss performance of classification model for POI recommendation, logistic regression and libFM. Experimental results shows unified POI recommendation framework CLW outperforms other approaches.