iGiST:一个改进的通用搜索树  

iGiST: An improved generalized search tree

作　　者：许向阳[1] 刘少治[1] 金光[1] 

机构地区：[1]华中科技大学计算机科学与技术学院

出　　处：《华中科技大学学报（自然科学版）》2002年第11期16-18,共3页Journal of Huazhong University of Science and Technology(Natural Science Edition)

摘　　要：分析了通用搜索树建树方法的不足以及结点溢出处理和下溢处理策略的缺陷 ,提出了一个改进的通用搜索树 ,增加了装树功能 ,提高了建树速度 ,同时对结点溢出处理和下溢处理策略做了改进 ,推迟了结点分裂和合并的时间 ,改进后的通用搜索树比原结构在系统效率和空间利用率方面有较大提高 .On the basis of analyzing disadvantages of GiST′s method to build the tree and the defection of strategies for handling node overflowing and node underflowing, an improved GiST was put forward. It has the method of packing tree, which increases the speed of building the tree. Some modifications such as handling node overflowing and node underflowing were proposed deferring the time of node splitting and merging. iGiST is better than GiST in system efficiency and space utilization ratio. Experiment results on a spatial database system were presented.

关 键 词：iGiST 可扩展性索引结构 通用搜索树 空间利用率 数据库系统 建树速度 溢出处理 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]