一种基于约简概念格的关联规则快速求解算法

关联规则挖掘是数据挖掘领域中重要的研究分支,已形成了较多的研究成果。然而,大多数基于频繁项集求解关联规则的挖掘算法需要多次扫描数据库。该文提出了一种基于概念格的关联规则快速求解算法,该算法仅需一次扫描数据库即可确定所有的频繁项集并且进而能够快速求解出关联规则。文章首先讨论了约简概念格(RECL)的构造原理,并详细描述了基于RECL的关联规则的挖掘算法,最后以实验证明了算法的正确性和优越性。

A Fast RECL-Based Algorithm For Mining Association Rules

Association rule mining is an important research branch of Data mining.However,most algorithms based on the frequent itemsets have to scan databases many times.This paper presents an algorithm that can find association rules quickly.The algorithm is based on the concept lattice and can certify all frequent temsets and all association rules with only one scan of databases.To describe this algorithm,this paper introduces the method to build the concept lattice with intensions reduced (RECL),discusses the fast RECL-based algorithm for mining association rules and analyses the experiments that have proved the efficiency and advantages of this algorithm.