Semantic-aware Spatial Keyword Querie

In recent years,with the fast development of GPS technologies and the widespread use of smart mobile devices,the Location Based Services(LBS)become more and more popular.For example,Baidu map turns into the most used tool in people's daily life when they want to go out for a traveling,which makes it more convenient.In LBS system,spatial keyword query is a very crucial core technique.Existing spatial keyword query methodologies mainly focus on the spatial and textual similarities,while leaving the semantic understanding of keywords in spatial web objects and queries to be ignored.To address this issue,this paper studies the problem of semantic based spatial keyword querying,which incorporates the semantic analysis of spatial web objects embedded in the query modeling and processing so that we can provide users more accurate and meaningful recommendations.It seeks to return the k objects most similar to the query,subject to not only their spatial and textual properties,but also the coherence of their semantic meanings.To support efficient query processing,we propose novel indexing structures,i.e.NIQtree and LHQ-tree,which integrate spatial,textual and semantic information in a hierarchical manner,so as to prune the search space effectively in query processing.NIQ-tree utilizes i Distance to index high dimensional topic distributions,which can help us locate the objects similar in semantic fast and accurately;while LHQ-tree hash the objects and the query into buckets and retrieves the objects in the same bucket more effectively.Extensive experiments on real datasets demonstrate the efficiency and scalability of the proposed index structure and search algorithms.At last,we implement a prototype system based on the semantic-aware spatial keyword queries.