基于条件函数依赖的挖掘算法研究

摘要

由于采用函数依赖( Functional Dependency,FD)对数据库的检测和修复还不够充分,现提出了条件函数依赖( Con-ditional Functional Dependency,CFD),其是在FD的基础上加入了语义约束。条件函数依赖的挖掘是一种重要的数据库分析技术,CFD挖掘是在FD挖掘的基础上通过条件分析进行更细粒度的信息挖掘,其时间复杂度较高。文中主要介绍了CFD的相关概念及CFD经典挖掘算法之一—CTANE,并对该算法效率进行改进。改进后的算法不仅可以提高数据挖掘过程中操作的效率,同时也将节省数据的存储空间。%Because the detection and repair of the database is not sufficient by Functional Dependency ( FD) ,Conditional Functional De-pendency ( CFD) is proposed,which is an extension of FD adding semantic constraints. The discovery of CFD is an important database a-nalysis technique,CFD mining do the more fine-grained information mines which based on FD mining,so the time complexity of CFD mining is higher than the latter. Introduce the related concept of CFD and one of the CFD classical mining algorithm—CTANE in this pa-per,and improve the efficiency of this algorithm. The improved algorithm can not only enhance the operating efficiency of the data min-ing process,but also save the data storage space.