Efficient Continuously Top-k Keyword Query for Moving Objects

With the growing of wireless communication and mobile applications, more and more users and contents are positioned by geographic locations. The importance of query objects involving both the location and textual proximity of content is highlighted. Existing studies mainly focus on keyword query for static objects or query for moving objects without text relevancy. In this paper, we address the challenge of continuously keyword query for moving objects in a server client model. As the objects are moving and the results are continuously refreshed by the query, communication costs will be huge due to frequently location updates. We employ the safe region techniques to reduce communication costs and a hybrid index based on TPR-tree [8] to index textual keywords of moving objects. Our major contributions are: (1) an efficient safe region generate solution along with the top-k calculation for circumstances that objects are moving and query lasts for a period of time, (2) efficient Safe region maintenance which significantly reduces communication costs and speed up result refreshment by cached results, and (3) lazy batch update strategy which reduces the safe region maintenance costs. The reduction of the communication cost and the efficiency of our solution are proved by experiments.