基于改进聚类分裂的动态R-树实现方法

An implementation method of dynamic R*-tree based on improved cluster splitting

在R*-树的构建过程中引入聚类技术能够有效地提高索引的性能,传统的k-means聚类算法对初始值非常敏感,聚类过程较为复杂。基于此,文中提出一种改进聚类分裂的动态R*-树实现方法,在节点分裂的过程中引进聚类技术,对R*-树的基本结构加以改进,从而获得动态的结构重组。实验表明,动态R*-树以略高的构建开销换取较高的查询效率,大幅度提高索引树的空间利用率,在批量数据动态加载和处理等方面具有较高的实用价值。

Clustering technologies can effectively improve the performance of R^＊- tree. The traditional K- means clustering algorithm is very sensitive to the initial value, and the process of clustering is very complex. Considering this, an improved clustering algorithm of the dynamic R^＊ - tree is proposed. In the process of node splitting, clustering technologies and improved structure can obtain dynamic structural reorganization of R^＊ - tree. Experimental results show that the dynamic R^＊- tree can achieve higher query efficiency and space utilization with a slightly higher construction cost, which has practical value in dynamic loading and processing of batch data.