数据仓库中维表模式的优化

OPTIMIZATION OF DIMENSION TABLE SCHEMA IN DATA WAREHOUSE

针对数据仓库使用雪花模式组织维表而带来的太多连接开销,以查询的时间开销和维表的存储开销作为衡量标准建立代价估算模型,设计基于遗传算法的维表模式优化算法。目的是实现维表模式自动调整,使系统在维表模式方面的维表存储开销和查询时间开销和最小。试验结果表明以较小的空间代价可以较大幅度地提高查询速度。

Lsing snowflake schema to group dimension table in a data warehouse will bring too much connection cost. To solve this problem, a cost estimation model is built up with the measurement standard of the time cost of queries and the storage cost of dimension tables,and the optimization algorithms of dimension tables schema is designed by using genetic algorithms. The purpose of the design is to realise automatie adjustment of the dimension tables schema so that the system has smallest storage cost of dimension tables and time cost of queries in dimension tables schema. From the experiment a conclusion can be drawn that the querying speed can be expedited remarkably at a smaller space cost.