一种有效的数据流最大频繁模式挖掘算法

针对频繁项集挖掘存在数据和模式冗余的问题,对数据流最大频繁项集挖掘算法进行了研究。针对目前典型的数据流最大频繁模式挖掘算法DSM-MFI存在消耗大量存储空间及执行效率低等问题,提出了一种挖掘数据流界标窗口内最大频繁项集的算法MMFI-DS,该算法首先采用SEFI-tree存储包含在不断增长的数据流中相关最大频繁项集的重要信息,同时删除SEFI-tree中大量不频繁项目,然后使用自顶向下和自底向上双向搜索策略挖掘界标窗口中一系列的最大频繁项集。理论分析与实验表明,该算法比DSM-MFI算法具有更高的效率,并能节省存储空间。

An efficient algorithm for mining maximal frequent itemsets over data streams

The paper focuses attention on the study of mining of maximal frequent itemsets from data streams to solve the problem of data and pattern redundance in frequent itemset mining, and in consideration of the problem of bad performance in operating time and memory space of the DSM-MFI, a typical algorithm for mining maximal frequent itemsets over data streams, presents an algorithm, called MMFI-DS. Firstly, the algorithm uses a new compressed tree, called the summary extended frequent item tree (SEFI-tree), to maintain the essential information about maximal frequent itemsets embedded in the stream so far, at the same time, a lot of infrequent items are deleted by pruning the tree. Then, it employs a top-bottom and bottom-top method to mine the set of all maximal frequent itemsets in landmark windows over the data stream. The theoretical analysis and experimental results show that the algorithm performs much better than the previous approaches.