一种条件函数依赖挖掘算法的分析与实现

摘要

条件函数依赖(Conditional Functional Dependence,CFD)的挖掘是一种重要的数据库分析技术,它是在函数依赖(Functional Dependence,FD)挖掘的基础上加入条件分析功能扩展而来,当前被应用于数据质量检测.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.文章详细介绍一种经典的CFD挖掘方法(即CTANE算法),对其实现方法提出了优化措施,并通过实验比较了不同实现方法的效率,得出了优化措施改进效率的结论.%The discovery of conditional functional dependence (CFD) from database is an important database analysis technique, which is an extension of functional dependence (FD) mining that adding semantic analysis,and currently used in data quality detection. CFD mining do the more fine-grained informations mines which based on FD mining,so the time complexity of CFD mining is higher than the latter. This paper introduces the details of a classic algorithm for CFD mining-CTANE, and proposes some optimized techniques for its implementation, and also compares the experimental results which illustrates the optimized approaches improve the efficiency.