A safe region based approach to moving K NN queries in obstructed space
Knowledge and Information Systems(2014)
摘要
The moving k nearest neighbor (M k NN) query has been studied extensively. Most of the studies assume no obstacle in the space. However, obstacles like rivers, buildings and private properties commonly exist in the space, and one may need to go around the obstacles to reach his/her nearest neighbors. In this paper, we study the moving k NN query in obstructed space with no predefined query object trajectory. We take a safe region based approach to solve this problem. In particular, we propose a method to compute a safe region w.r.t. a data object. In this safe region, the query object can move freely, while the data object is kept in the query object’s k NN set. By combining the safe regions of the data objects near the query object, we formulate an overall safe region where the query object’s k NN set keeps stable. We propose an algorithm based on the safe regions to process the moving k NN query in obstructed space. Extensive experiments show that the proposed algorithm significantly reduces the communication and the computation costs for query processing. Our algorithm outperforms a baseline algorithm by up to two orders of magnitude under various settings.
更多查看译文
关键词
Moving nearest neighbor query,Obstructed space,Safe region,Fixed-rank region
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络