摘要
通过对Apriori算法进行的分析与研究,发现其在实用中存在两个主要问题:生成的关联规则具有相当大的冗余性;有可能挖掘出一条支持度和置信度均很高,但却是无趣的、甚至是虚假的关联规则,且不能产生带有否定项的规则。鉴于此,该文给出了关联规则的两个性质和引入兴趣度的第三个度量———相关支持度,设计了挖掘出有效关联规则算法,较好地解决了上述问题。
By analyzing and studying Apriori algorithm,this paper finds two problems.Firstly,generated association rules are quite redundant.Secondly,it is probable to find an association rule,which possess high support and confidence,but is uninteresting,and even is false.Furthermore,a rule with negative-item can't be produced.Thus,this paper,after giving the two properties of association rules and introducing interest measure's third threshold—correlation support,effective association rules are worked out and able to solve above problems well.
出处
《计算机工程与应用》
CSCD
北大核心
2003年第19期196-198,共3页
Computer Engineering and Applications
基金
河南省自然科学基金项目(编号:111070600)