Top-k Spatio-textual Similarity Search

Location-based services have attracted significant attention for the ubiquitous smartphones equipped with GPS systems. These services (e.g., Twitter, Foursquare) generate large amounts of spatio-textual data which contain both geographical location and textual description. In this paper, we study a prevalent top-k spatio-textual similarity search problem: Given a set of objects and a user query, find k most relevant objects considering both spatial location and textual description. We make the following contributions: (1) We propose a TA-based framework and devise efficient algorithms to incrementally visit the objects with current highest spatial or textual similarity. (2) We explore a hybrid partition pattern by integrating spatial and textual pruning power. We further propose a partition-based algorithm which can significantly improve the performance. (3) We have conducted extensive experiments on real and synthetic datasets. Experimental results show that our methods outperform state-of-the-art algorithms and achieve high performance.