A Hierarchical Index Structure for Region-Aware Spatial Keyword Search with Edit Distance Constraint

Location-based services have become widely available on a variety of devices. Due to the errors in user input as well as geo-textual databases, supporting error-tolerant spatio-textual search becomes an important problem in the field of spatial keyword search. Edit distance is the most widely used metrics to capture typographical errors. However, existing techniques for spatio-textual similarity query mainly focused on the set based textual relevance, but they cannot work well for edit distance due to the lack of filter power, which would involve larger overhead of computing edit distance. In this paper, we propose a novel framework to solve the region aware top-k similarity search problem with edit distance constraint. We first propose a hierarchical index structure to capture signatures of both spatial and textual relevance. We then utilize the prefix filter techniques to support top-k similarity search on the index. We further propose an estimation based method and a greedy search algorithm to make full use of the filter power of the hierarchical index. Experimental results on real world POI datasets show that our method outperforms state-of-the-art methods by up to two orders of magnitude.