基于频繁项集挖掘最大频繁项集和频繁闭项集

Mining of maximum frequent itemsets and frequent closed itemsets based on frequent itemsets

提出了基于频繁项集的最大频繁项集(BFI-DMFI)和频繁闭项集挖掘算法(BFI-DCFI)。BFI-DMFI算法通过逐个检测频繁项集在其集合中是否存在超集确定该项集是不是最大频繁项集;BFI-DCFI算法则是通过挖掘所有支持度相等的频繁项集中的最大频繁项集组合生成频繁闭项集。该类算法的提出,为关联规则的精简提供了一种新的解决方法。

In this paper,a new kind of algorithms BFI-DMFI （Mining Maximum Frequent Itemsets） and BFI-DCFI（Mining Frequent Closed Itemsets） is proposed.In BFI-DMFI,we can confirm whether a frequent itemsets is also a maximum frequent itemsets through detecting whether exiting their superset itemsets in frequent itemsets.In BFI-DCFI,in order to generate frequent closed itemsets,we can make use of mining maximum frequent itemsets from frequent itemsets which have equal support.This kind of algorithms provides a new method for reducing the set of association rules.