关联规则挖掘Apriori算法的研究与改进  

The Research and Improvement of Apriori Algorithm for Mining Association Rules

作　　者：付沙[1] 周航军[1] 

机构地区：[1]湖南财政经济学院,湖南长沙410205

出　　处：《微电子学与计算机》2013年第9期110-114,共5页Microelectronics & Computer

基　　金：湖南省高等学校科学研究项目(12C0559)

摘　　要：针对传统Apriori算法的不足之处,提出两种Apriori改进算法,分别基于构造辅助表和项集求交集策略.改进算法大幅度减少扫描数据库的次数,缩减对不必要事务的扫描时间,显著提高频繁项集的生成效率,从而使算法达到更高的运算效率.实验结果表明,两种改进算法都是有效的关联规则挖掘方法,且比Apriori算法性能更优,亦为关联规则挖掘研究提供了一些科学可行的新思路.This paper proposed two Apriori improved algorithm with the inadequacies of the traditional Apriori algorithm, respectively based on the constructed auxiliary table and itemsets intersection strategies. Improved algorithm significantly reduced the number of times of scanning the database, and reduced unnecessary transaction scan time~ improve the generation efficiency of frequent itemsets significantly, so that the algorithm can achieve higher operation efficiency. The experimental results show that the two improved algorithms are effective association rule mining method, and is better than the Apriori algorithm performance, and also provides a feasible new ideas to the association rule mining research.

关 键 词：数据挖掘 关联规则 Apriori辅助表 交集策略 频繁项集 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]