ETL过程的思考  

Thinking in ETL Process

作　　者：王亮[1] 葛玮[1] 

机构地区：[1]西北大学软件工程研究所,陕西西安710127

出　　处：《计算机技术与发展》2008年第10期130-132,共3页Computer Technology and Development

基　　金：国家"863"计划资助项目(2004AA115090)

摘　　要：ETL是构建数据仓库的一个非常重要的环节,可以这样认为:ETL就是整个数据仓库系统乃至整个决策支持系统的基石。如何设计高效的ETL过程就成为了众多计划或正在实施数据仓库项目的企业考虑的重要问题。从前期的数据理解阶段入手,分别讨论了数据的抽取、清洗转换、装载等不同阶段需要考虑的设计问题及相应的解决方案。提出了以数据理解为根基,以清洗转换为中心的设计思想,并给出了具体的实施步骤。ETL is one of the major processes when building a data warehouse. Can consider that ETL is the basis of the data warehouse even though the whole decision support system. Many enterprises which is planning or beginning to build a data warhorse now is highly concerned about how to design the ETL process effectively. Starts with the data comprehension, and discusses the design issues and its solutions about data extraction, data cleaning and data loading. Propose a method which is based on data comprehension and centered on data cleaning to design the ETL process and describe the steps to follow.

关 键 词：ETL 商业智能 数据仓库 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]