Main Article Content
Abstract
Predicting the popularity of Point of Interest (POI) has become gradually more important for location -based services, like POI recommendation. Already existing system only just achieve a suitable concert by reason of the shortage of POI’s information that tendentiously confines the recommendation to popular scene spots, and ignores the unpopular attractions with potentially precious values. In this paper, we put forward a novel approach, termed Hierarchical Multi-Clue Fusion (HMCF), for predicting the popularity of POIs. Particularly, in order to deal with the problem of data sparsity, we propose mostly to describe POI using various types of user generated content (UGC) (e.g., text and image) from multiple sources. Then, we invent an effective POI modeling technique in a hierarchical manner, which all together injects semantic knowledge as well as multi - clue representative power into POIs.