Mining Frequent Weighted Itemsets Using Extended N-List and Subsume

Discovering frequent weighted itemsets(FWIs) is an important research in practical applications of field of data mining. Recently the PrePost algorithm based on the idea of N-lists has been presented. In this paper, we propose an improved version method ENSFWI(extended N-list subsume-based algorithm for finding FWIs). The subsume concept and related theorems are proposed to calculate the weighted supports of itemsets fast and generate directly FWIs without extended N-list intersection, and then an algorithm is built based on these concept for efficiently mining FWIs. It is shown by experimental results that our approach not only results in shorter execution times, but also reduces the memory usage when run on very large and dense database.