基于集体度一置信度的关联规则挖掘

Mining Association Rules Based on the Framework of Collective and Confidence

总结并研究了基于集体度-置信度的关联规则挖掘算法,用集体度代替支持度对搜索空间进行压缩,成功地解决了传统的频繁关联规则挖掘存在的属性集产生上的欺骗性及处理稠密数据集方面的缺陷.

The authors summarize and introduce some achievement in the study of mining association rule in recent years, such as mining association rules based on the framework of collective and confidence. The algorithm overcomes some inadequacies that traditional algorithms have, such as: spuriousness in itemset generation including phenomenon of negatively association, and the inefficiency in dealing with dense data sets.