A generic approach for indexing Top-k spatial keyword queries

Abstract

A top-k spatial keyword query returns the k best spatio-textual objects ranked based on their proximity to the query location and relevance to the query keywords. It allows users to find the objects in the order of importance according to the users` preference on spatial proximity and textual relevancy. In this dissertation, we propose a generic approach for indexing top-k spatial keyword queries. The goal of the dissertation is to exhaustively investigate possible indexing methods for top-k spatial keyword queries and to identify the most appropriate method for a certain usage pattern (i.e., a set of query loads and frequencies). To achieve this, we model an indexing method for top-k spatial keyword queries as a combination of various clustering techniques. First, we propose a new method for top-k spatial keyword queries employing a novel clustering technique. Second, we propose an index generation model for top-k spatial keyword queries by combining multiple clustering techniques.

In the first part of this dissertation, we propose a new method for the separate index approach called Rank-Aware Separate Index Method(RASIM) for top-k spatial keyword queries. Possible indexing approaches for top-k spatial keyword queries can be classified into two categories: the separate index approach and the hybrid index approach. The separate index approach maintains the spatial index and the text index independently and can accommodate new data types. However, it has been considered difficult to support top-k pruning in the separate index approach since it requires clustering/ordering the objects by two different criteria---one for top-k pruning and the other for efficient merging. Naturally, all the existing methods have been based on the hybrid index approach. Thus, we need to develop a new method supporting top-k pruning based on the separate index approach by devising a technique that clusters the objects by multiple criteria. RASIM supports both top-k pruning and ...