Social-Aware Top-k Spatial Keyword Search

The boom of the spatial web has enabled spatial keyword queries that take a user location and multiple search keywords as arguments and return the objects that are spatially and textually relevant to these arguments. Recently, utilizing social data to improve search results, normally by giving a higher rank to the content generated or consumed by the searcher's friends in the social network, has been studied in the information retrieval (IR) community. However, little attention has been drawn to the integration of social factors into spatial keyword query processing. In this paper, we propose a novel spatial keyword query, Social-aware top-k Spatial Keyword (SkSK) query, which enriches the semantics of the conventional spatial keyword query by introducing a new social relevance attribute. A hybrid index structure, called Social Network-aware IR-tree (SNIR-tree), is proposed for the processing of SkSK queries. To further improve the query response time, an x-hop localized algorithm is developed. Empirical results demonstrate that the proposed index and algorithms are capable of excellent performance.published_or_final_versio