An adaptive point-of-interest recommendation method for location-based social networks based on user activity and spatial features

Point-of-interest (POI) recommendations can help users effectively explore new locations according to their preferences, which is an important research aspect for location-based social networks (LBSNs). However, most existing POI recommendation methods lack adaptability when making recommendations for users with different preferences, which causes unsatisfactory recommendation results. To this end, in this paper, we propose an adaptive POI recommendation method by combining user activity and spatial features, which can operate adaptively according to user activity. First, we extract three-dimensional user activity, time-based POI popularity and distance features using a probabilistic statistical analysis method from historical check-in datasets on LBSNs. Second, we devise a user activity clustering algorithm that is based on fuzzy c-means and compute POI popularity by applying smoothing technology to adjacent continuous time slots. Finally, we propose an adaptive recommendation scheme, which includes a two-dimensional Gaussian kernel density estimation algorithm and a one-dimensional power-law function algorithm with POI popularity according to user activity. Extensive experiments on Foursquare and Gowalla datasets show that the proposed method outperforms the baseline methods in terms of both precision and recall.