改进的半连接查询优化算法  

An improved algorithm with semi-join query optimization

作　　者：钱磊[1] 于洪涛 

机构地区：[1]燕山大学信息科学与工程学院,河北秦皇岛066004

出　　处：《燕山大学学报》2012年第2期178-182,共5页Journal of Yanshan University

基　　金：河北省自然科学基金资助项目(F2011203219)

摘　　要：在多关系连接查询中,普通半连接查询方法没有优化子查询的半连接顺序,导致查询代价较高,为此,本文提出了一种改进的半连接查询优化算法。首先,将多关系连接组织成较小代价的类树形结构,然后利用半连接操作对处于根节点处的关系进行最大化缩减,并以此为基础利用PERF位向量对其他节点进行缩减,最后回收缩减关系进行连接操作。模拟实验表明改进的半连接查询优化算法能够有效地缩减查询关系,降低查询代价。In multi-join query, the ordinary semi-join query method does not optimize the order of sub-query＇s semi-join, which results in high query cost. An improved semi-join query optimization algorithm is proposed to solve this question in this paper. First, the connections of relations are tissued to small cost simulated tree structure. Second, the relation is reduced which is at the root node of simulated tree structure absolutely, then other nodes are reduced by using PERF bit vector based on root node. Finally, the reduced relations are recycled to join together. The simulated experiment shows that the evolutionary semi-join query optimi- zation algorithm can reduce the query relations, lower the query cost effectively.

关 键 词：多关系连接查询优化 半连接 类树形结构 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]