基于二进制编码的优化关联规则挖掘算法  

On Mining Algorithm of Optimal Association Rules Based on Binary Arranging

作　　者：王仕平[1] 

机构地区：[1]成都学院信息科学与技术学院,成都610106

出　　处：《西南师范大学学报（自然科学版）》2013年第9期44-49,共6页Journal of Southwest China Normal University(Natural Science Edition)

摘　　要：提出了一种基于二进制编码的优化关联规则挖掘算法,该算法是按项目支持数的升序从高到低地编制二进制位,然后将事务转换成数字事务,通过构建候选数字事务区间来搜索频繁数字事务,最后产生关联规则.该算法的原理简单,减少了冗余候选项和扫描次数;实验结果表明该算法比同类挖掘算法更快速而有效.At present in the research of mining algorithm of association rules, these mining algorithms based on binary are obviously much better than the algorithm based on Apriori, but they are still of redun dant candidates and eomputing. And then an mining algorithm of optimal association rules based on binary arranging has been proposed in this paper, with binary digits from highness to lowness arranged according to ascending of support for each item, transaction into digital transaction transformed, frequent digital transaction discovered via establishing candidate digital transaction section, and, finally, association rules generated. The principle of algorithm is simple, which makes redundant candidates and scanning times re duced. The experimental results indicate that the algorithm is faster and more efficient than similar mining algorithms.

关 键 词：关联规则 二进制项目 数字事务 候选数字事务区间 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]