一种决策树算法在数据仓库ETL中的应用与研究  

Application and Research of Decision Tree in Data Warehouse's ETL Process

作　　者：高小飞[1] 冯杰 

机构地区：[1]上海电力学院电力与自动化学院,上海200090 [2]上海亚软信息技术有限公司,上海201204

出　　处：《计算机与现代化》2009年第4期65-67,70,共4页Computer and Modernization

基　　金：上海高校选拔培养优秀青年教师科研专项基金资助项目(Z-2006-81)

摘　　要：ETL是建立企业级数据仓库的核心环节,传统ETL受CPU、I/O等系统资源的影响较大,从而影响整个数据仓库运转性能。本文通过建立专家知识库,并在此基础上引入决策树算法,从而动态地调整ETL策略,并取得良好的实践效果,该架构具有一定实用价值。ETL is a core process of building a data warehouse. System resources like CPU and I/O have great effects on the process of ETL so that the performance of DW is decreased. Through building expert knowledge base and using decision tree algorithm, the ETL architecture is dynamically adjusted so that satisfactory effect is obtained. The ETL architecture is of practical use.

关 键 词：数据仓库 决策树 ETL 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]