基于FP-Tree的最大频繁项目集挖掘算法

摘要

Mining maximum frequent itemset is a key problem in mining association rules.Candidate set generation is costly based on Apriori,especially when there exist prolific pattern or long patterns.In this paper,an algorithm for mining maximum frequent itemsets based on FP-Tree is proposed.The algorithm MFIP-Miner(maximum frequent itemsets pattern miner)without candidate generation in the mining period,therefore it increases the mining efficiency.%挖掘最大频繁项目集是关联规则挖掘中的关键问题,基于Apriori产生候选项目集需要付出很高的代价,尤其是在存在大量强模式或长模式的时候.提出一种基于频繁模式树(FP-Tree)的最大频繁项目集挖掘算 MFIP-Miner (maximum frequent itemsets pattern miner),其挖掘过程无需产生候选项集,从而提高挖掘效率.