基于项集的多支持度关联规则挖掘算法  

Itemset-based Multiple Min-Support Association Rule Mining Algorithm

作　　者：余晓平[1] 刘丽娅[1] 朱东芹[1] 

机构地区：[1]石河子大学计算机科学与技术学院

出　　处：《微计算机信息》2009年第33期147-148,112,共3页Control & Automation

摘　　要：提出利用预存储的信息挖掘基于项集的多最小支持度关联规则挖掘算法,允许用户指定多个最小支持度,在较短的时间内找出所关心的规则,从而解决了固定支持度挖掘算法中当最小支持度设置太低或太高时出现的组合爆炸和低频率规则不被发现的问题。理论证明了相关性质,并实验验证了算法的有效性。A new itemset-based multiple minimum supports using prestored information mining association rules algorithm is proposed. The algorithm allows the user to specify multiple minimum supports and find interesting rules in short time. It resolves the problem in single support mining algorithm that is when minimum support is set too low, it will cause combinatorial explosion and when minimum support is set too high, rare items will not be found. Relative properties are proved by theory and experimental result show its efficiency.

关 键 词：频繁项集 多支持度 关联规则 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]