关联规则挖掘算法分析与探讨  

Analysis and Discussion on Association Rule Mining Algorithms

作　　者：刘晓玲[1] 郭龙[1] 

机构地区：[1]济南职业学院计算机系,山东济南250103

出　　处：《软件导刊》2013年第7期66-68,共3页Software Guide

摘　　要：关联规则挖掘算法是数据挖掘领域的主要研究方向之一。对几种经典的关联规则挖掘算法进行了分析、探讨和比较,给出了一种基于支持矩阵的、不需要产生候选项目集的算法设计思想。算法为事务数据库中的每个项目设置二进制向量,利用逻辑与运算构造支持矩阵来挖掘频繁项目集,极大地节省了存储空间,提高了算法运行效率。Association rule mining algorithm is a major research direction of data mining.In this paper,several classical algorithms for mining association rule have detailed analysis,discussion and comparison.A design thought of algorithm is proposed which based on support matrix,and do not need to produce candidate itemset.In this algorithm,binary vector has been set for each database item,using logic and operation to structure support matrix and to mining frequent itemsets.This algorithm saves storage space greatly and improves the efficiency of the algorithm.

关 键 词：关联规则 候选项目集 频繁项目集 支持度 

分 类 号：TP312[自动化与计算机技术—计算机软件与理论]