基于改进频繁模式树的最大频繁项目集更新挖掘算法

在挖掘最大频繁项目集的过程中,通过改变最小支持度阈值可以挖掘更有用的最大频繁项目集,为此提出了一种最大频繁项目集更新挖掘算法UAMMFI(Updating Algorithm for Mining Maximal Frequent Itemsets)。算法基于改进后的频繁模式树结构,在更新挖掘过程中,不需产生候选项目集和条件模式树,并且充分利用先前已挖掘的最大频繁项目集中包含的信息,快速更新挖掘出最小支持度阈值变化后的最大频繁项目集。实验结果表明,算法能够高效更新挖掘最大频繁项目集。

AN UPDATING ALGORITHM FOR MINING MAXIMAL FREQUENT ITEM SETS BASED ON IMPROVED FREQUENT PATTERN TREE

In mining the maximal frequent item sets, we can get more useful maximal frequent item sets by changing the minimum support threshold. Furthermore, we proposed a new algorithm UAMMFI(updating algorithm for mining maximal frequent item sets) for mining maximal frequent item sets for the demand. The algorithm didn''t generate candidate item sets and construct conditional pattern tree based on improved frequent tree. We also make full use of the information contained in maximal frequent item sets previously mined in the process of updating mining maximal frequent item sets. Finally, we can efficient mining new maximal frequent item sets after the minimum support threshold change. Based on experiment results, we can see that the algorithm has excellent performance.