一种基于Rymon枚举树的快速挖掘无关集算法

摘要

In view that traditional mining δ free sets algorithm exist generating candidate itemsets,traversing all direct subsets,and so on leading to low efficiency,a free sets determine lemma and a fast mining free sets algorithm FMFS were proposed.FMFS used Rymon setenumeration tree as searching space and combined with pruning strategy,and then used the free sets determine lemma to quickly determine the candidate itemset.Experimental results showed FMFS algorithm not only mined all free sets,but also showed better performance than existed mining algorithm.%针对传统的挖掘算法在挖掘δ无关集时存在重复生成候选项集、遍历子项集等导致挖掘效率过低的问题,提出一个无关集判断定律,进而给出一种快速挖掘无关集算法FMFS.该算法利用Rymon枚举树作为搜索空间,结合一定的剪枝策略,再利用这个无关集判断定律对候选项集进行快速筛选.实验结果表明,该算法不仅能够挖掘出所有的无关集,且挖掘过程中的时间消耗优于目前已有算法.