Research On Near Neighbor Query Of Hybrid Data In Spatial Database

In recent years,near neighbor query technology has become more and more important in a variety of applications.The existing near neighbor query methods mainly abstract spatial objects as points or segments.However,in practical applications,only abstracting spatial objects as points or line segments will affect the accuracy and efficiency of query to a great extent,so sometimes it is necessary to abstract spatial objects as hybrid data of points and line segments according to the actual situation.Because the existing near neighbor query methods can not directly and effectively solve the near neighbor query problem of hybrid data,this paper proposes the k-nearest neighbor query method of hybrid data in spatial database and the group nearest neighbor query method of hybrid data in spatial database.Firstly,this paper takes hybrid data as the research object,extends the traditional point based Voronoi diagram and line based Voronoi diagram,and proposes the concept and properties of hybrid data Voronoi diagram.Compared with traditional Voronoi diagram,the edges of hybrid Voronoi diagram include point to point,point to line,distance bisector of line to line.Therefore,this paper describes the construction of hybrid Voronoi edges in detail.Then,this paper proposes a hybrid data k nearest neighbor query algorithm,that is to get the k data objects closest to the target object.The pruning algorithm proposed in this paper can effectively remove the data objects that can not be the query results,and get the candidate set of k-nearest neighbor query results;in the data set refining stage,according to the position relationship between different objects,the specific distance calculation method is given according to the distance between the object and the query object,the k-nearest neighbor query results of hybrid data are obtained.Finally,the efficiency and accuracy of the proposed algorithm are proved by theoretical research and experimental results.Finally,this paper proposes a hybrid data group nearest neighbor query algorithm,that is to get the data object with the minimum sum of distances from all query objects.This algorithm extends the nearest neighbor query when the number of query objects is more than 1.Firstly,the hybrid Voronoi diagram is constructed;secondly,the candidate set is obtained by pruning the data set in the case that the query objects can form a hybrid convex hull and the query objects can not form a convex hull;then the sum of the distances between the data objects in the candidate set and each query object is compared to get the correct query result;finally,theoretical research and experimental results prove the efficiency and accuracy of the proposed algorithm.