数据仓库及其设计规范化

在严格数据仓库概念基础上，说明了数据仓库与多维数据库之间关系及数据仓库的特性和优势，重点讨论并给出一种规范化的数据仓库设计和粒度模型设计方法，粒度设计在数据仓库建模中十分重要，实质上，它在数据仓库模型设计的一些重要指标间提供了一项交换技术，以保证数据仓库系统结构稳定、存储与访问性能良好，但粒度设计往往凭经验调试完成，根据专家系统技术，提出一种动态粒度划分方法，很好地解决了数据仓库的粒度设计问题。

Normalized designing methods for data warehousing

Based on the strict concept of data warehouse, the relations between data warehouse and multi dimensional database and the performances of data warehouse are explained. The normalized methods for data warehouse and data granularity model designing are given and discussed. Granularity model designing is very important in data warehousing. To make structure steady, the performance of data store and access well, a kind of trade off technique is given among these important factors of data warehousing essentially. Nevertheless, granularity model designing is completed generally according to expert experiences. According to the techniques of expert system, a dynamic method of granularity model designing is given to solve the problem well.