基于关系数据库的数据仓库星形模式下维使用原则的研究

Research of restricted use of dimensionality in star schema of data warehousing based on RDS

星形模式是基于关系数据库的数据仓库中的一个著名概念,由于星形连接模式的设计思想能够满足人们从不同观察角度(维)分析数据的需求,所以在基于关系数据库的数据仓库的设计中广泛地使用了星形模式。在使用数据仓库来回答综合性问题的场合,通常可以使用OLAP工具实现记录不多的较高粒度表的维度旋转来满足不同分析的需要;而在数据仓库中较高粒度表中记录非常多或者还要经常回答细节问题的场合,则还必须对数据仓库中记录非常多的较高粒度的表或者细节级表进行维度转换。但通常的OLAP工具难以处理几十万条记录数据表的维度旋转,针对这种应用场合,提出了一种有选择地使用维的星形模式,在事实表中避开使用要旋转的维,用存贮过程编写程序高效地实现OLAP工具相应的功能,对星形模式下维的使用原则做出了一定的探索。

Star schema is a famous conception in data warehousing based on RDS. It's widely used in data warehousing based on RDS for its convenience for people to analyze data from different angle. In the situation for people using data warehousing to answer the all-around question, OLAP tools is usually used to circumrotate the dimensionality of high granularity tables with a few records for the requirement of analysis. If there is a great deal records in high granularity tables or In the situation for people using data warehousing to answer the detail question, it's necessary to circumrotate the dimensionality of those high granularity tables or the detail tables in data warehousing. While it's very difficult for OLAP tools to circumrotate the dimensionality of tables with more than ten thousands records, a new star schema with restricted is issued using of dimensionality. In this star schema the fact tables without the dimensionality to be circumrotated is designed, and then a program is developed using stored procedure to implement the corresponding function of OLAP tools in high efficiency. In this process, definite research is done about the rule of using dimensionality in star schema in RDS.