Geo-Social Keyword Top-k Data Monitoring over Sliding Window

摘要

Recently, in many applications, points of interest (Pols) have generated data objects based on Publish/Subscribe (Pub/Sub) model, and users receive their preferable data objects. Due to the prevalence of location based services and social network services, in addition, locations, keywords, and social relationships are considered to be meaningful for data retrieval. In this paper, we address the problem of monitoring top-k most relevant data objects over a sliding window, w.r.t. distance, keyword, and social relationship. If we have a lot of queries, it is time-consuming to check the result update of all queries. To solve this problem, we propose an algorithm that maintains queries with a Quadtree and accesses only queries with possibilities that a generated data object becomes top-k data. Moreover, we utilize k-skyband technique to quickly update query results. Our experiments using real datasets verify the efficiency of our algorithm.