快速挖掘最大频繁项集

A Fast Algorithm for Mining Maximal Frequent Itemsets

提出了一种基于布尔矩阵的最大频繁项集挖掘算法,通过将FP-tree映射成布尔矩阵和权值表,运用布尔逻辑运算进行矩阵投影操作得到最大频繁项集,算法在挖掘过程中不用生成最大频繁候选项集,从而大大提高了算法的时间效率和空间可伸缩性。

This paper proposes a maximal frequent itemsets mining algorithm based on boolean array. This algorithm projects FP-tree into a boolean array and into weight table, and obtains maximal frequent itemsets using boolean logic operation.This algorithm needn＇t generate maximal frequent candidate itemsets in mining process, therefore greatly improves the mining efficiency in time and space scalability.