摘要
针对传统基于支持度-置信度桩架关联规则挖掘算法生成规则只考虑支持度和置信度的因素、没有考虑规则两者本身内在的关系并因此产生大量无效的规则的情况,提出了一种采用相关性精简关联规则产生的算法,此算法增加了一个度量——相关性来精简关联规则的产生。实验表明该算法在继承传统算法的优点的同时在一定的程度上提高了规则的有效性,降低规则冗余。
The rules from the traditional association discovery algorithm concern only the measure of support and confidence index without considering of the correlation of the cause and result of the rule. Lots of rules produced are invalid in practice. An algorithm from reduced correlation rule is presented based on the correlation analysis, it uses correlation to reduce the generating of association rules. The experiment shows that this algorithm can improve the validity and reduce the redundancy of rules with remaining the advantages of the traditional association discovery algorithm.
出处
《江苏科技大学学报（自然科学版）》
CAS
北大核心
2007年第1期56-60,共5页
Journal of Jiangsu University of Science and Technology:Natural Science Edition
基金
省自然科学基金(BK2004058)
关键词
数据挖掘
关联规则
有效性
相关度
datamining
association rules
validity
correlation
