构建数据仓库模型的一种方法

Study on building data warehouse model

构建了具有 4个存储结构和 3个数据流的数据仓库模型 ,并以参与设计的数据仓库系统为例 ,论述了数据仓库模型的构建过程。较之其他数据仓库系统 ,本系统创建了第二索引 ,即一个高性能的存储 ,以空间换时间 ,既实现了企业数据集成 ,又提高了系统的查询速度 ;在数据仓库中设置了代理键 ,简化了多数据源数据的集成工作 ,提高了数据抽取、转换和加载的速度 ,并为数据仓库系统的迭代式开发提供了有利条件。提出了构建数据仓库模型的建议 :采用星系模型 ,原子级数据模型和汇总级数据模型并存的形式 ,设立代理键。

Being with taking the background of data warehouse for instance, the process of constructing the data warehouse model was discussed in detail, and a kind of data warehouse architecture containing four store-structures and three data-flows was put forward. Two new ideas was given compared with other similar systems, firstly, the second index for data warehouse was created, for the purpose of accomplishing the enterprise data integration and system query speed, which win the time at the cost of space. Secondly, the surrogate key was set in the data warehouse, which makes the integration of multiple data source simple, improves the speed of data draw-out, data conversion, data load and makes it possible for the iterative development to the data warehouse system. An advice of constructing the data warehouse model was put forward: adopting star model, merging atom data model and statistics data model, seting the surrogate key.