无损BCNF分解算法的改进

Improvement of BCNF Lossless Decomposition Algorithm

在关系数据理论中,为了优化数据库的设计,要求将低一级的关系模式转换为高一级的关系模式集.根据关系数据理论,任何一个关系模式都能转换为无损的BCNF关系模式集,但不一定能保函.为了使得转换后得到无损保函的BCNF模式集,本文引入图论中的有向图来表示函数依赖关系,并将有向图结合运用到关系模式的规范化过程中,使得模式分解能得到无损保函的BCNF模式集,从而达到改进传统的BCNF分解算法,优化分解的效果.

In relational data theory,it is required for optimizing database design that a lower-level relation schema be converted into a higher level relation scheme set. According to relational data theory,any relation schema can be converted to BCNF lossless relation scheme set,but not necessarily preserving function dependency. To obtain lossless and function preserved BCNF schema after the conversion,this paper introduces directed graph to express function dependencies. Directed graph is also used in the normalization of relational schema,to make sure the decomposition of schema obtain lossless and function preserved BCNF schema set,to improve the traditional BCNF composition algorithm.