一种结合空间聚类算法的R树优化算法  

Optimization algorithm for R-tree combining with spatial-clusting

作　　者：汪璟玢[1] 

机构地区：[1]福州大学数学与计算机科学学院,福州350108

出　　处：《计算机工程与应用》2014年第5期112-115,共4页Computer Engineering and Applications

基　　金：空间数据挖掘与信息共享教育部重点实验室2009年开放基金(No.201006)

摘　　要：研究R树特点,考虑了离群点对R树结点构造的影响,结合改进的k-medoids聚类算法提出了一种新的R树构造算法。与传统R树相比,新算法下构造的R树结点更加紧凑。通过实验证明,该优化算法构造的R树在查询性能方面的改进是明显的。This paper proposes a new method of constructing R-tree by studying its features, considering the outliers’ influence on constructing R-tree’s nodes and combining with improved k-medoids clustering algorithm. The R-tree based on this method has more compact structure compared with traditional R-tree. The results of the study show that, due to the optimization of structure, the improvement of R-tree in query efficiency is effectively.

关 键 词：R树 空间聚类 空间数据 地理信息系统(GIS) GEOGRAPHIC INFORMATION System(GIS) 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]