为物化视图选择构造搜索空间的新策略IMVPP  

New Strategy of Constructing Search Spaces for Materialized IMVPP

作　　者：夏小玲[1] 张红[1] 

机构地区：[1]东华大学计算机科学与技术学院,上海201620

出　　处：《计算机科学与探索》2010年第5期473-479,共7页Journal of Frontiers of Computer Science and Technology

基　　金：上海市科委公关项目~~

摘　　要：数据仓库中物化视图选择算法的代价与搜索空间的尺寸紧密相关。提出了一种基于输入查询的公共子表达式的候选视图搜索空间构造方法IMVPP,利用算法1计算出的公共子表达式,能被其他查询共享,并可对输入查询进行重写,有利于缩减视图搜索空间,提高查询效率。理论分析与实验结果表明,此方法是有效、可行的。The materialized view selection algorithms are closely related to the size of the search space. This paper proposes a new strategy IMVPP of constructing search space for materialized views selection based on the common expression of the input queries,which calculate the max common sub-expression using algorithm 1,can be shared by the other queries,and can rewrite the input queries,which is of benefit to reduce the search space and improve the querying efficiency. Theoretical analysis and comparative experiment indicate that IMVPP operates more effectively than MVPP.

关 键 词：数据仓库 物化视图 物化视图处理方法 查询图 公共子表达式 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]