A Research On Methods Of Approximate String Reverse Furthest Neighbors Query

With the development of location-based services, many location information and keyword information related to the position are needed together to provide users with a variety of customized services. For example, We can use a mobile phone query the characteristics of restaurant closed to a location, or query what place can provide conform to user requirements of rental housing, etc. At present the demand of these applications is rising, the study of such problems has become an important scientific significance and application value of the research subject. Spatial database query and processing is widely and deeply researched in the field of database. There are many applications of spatial query type in real life like nearest neighbor query and reverse nearest neighbor query has been widely research, and many relative classic algorithms have been proposed.Based on large scale space and text data as the research object, we research spatial reversed furthest neighbor query and process based on approximation keyword, efficient index structure and query algorithm is proposed. This paper first defines the approximate string reverse furthest neighbors query problem, this kind of query indicates that search for all points that take a given query point as its farthest neighbor while satisfy the similarity conditions of the target text in a spatial data set. From the existing literature, the reverse furthest neighbors query focus on spatial location information, does not include the approximate query text. We propose two basic algorithms and an efficient algorithm based on index structure of Filter-Rtree to solve this problem the approximate string reverse furthest neighbors query based on the existing spatial reverse furthest neighbors query algorithm and approximate keyword query algorithm. Through experimental tests on real data sets, we demonstrate the effectiveness of these algorithms, we also analysis and compare their performance and efficiency.