Survey on Spatial Item Recommender System using SPORE Model

Abstract: In recent year, Location based social network provide different user location features and recommendation. In location recommendation provide geographical influence and social influence with spatial item, and temporal data. A newly features stored user visit location. The user-item matrix is very sparse, which creates a big challenge for traditional collaborative filtering-based recommender systems. In Location based service has low sampling rate which existing prediction rate. In proposed system, we implement sequential personalized spatial item recommendation framework which provide user interest personal influence. We calculate the performance of SPORE on two datasets (Foursquare, Twitter) and one large-scale synthetic dataset. The results improvement SPORE ability to recommend spatial items, in terms of both efficiency and effectiveness, compared with the state-of-the-art methods.

Keywords: Location-based service, Location-based Social Network, Temporal data, spatial item.