Một phương pháp mới khai thác các mẫu hữu ích cao thú vị

Abstract

Mining high utility itemsets is an important problem in datamining. This supplies more benefit itemsets than the traditional itemsets, but many high utility itemsets is not important because the poor relations among the items inside them. Recently, the theory of the high utility and interesting itemsets and the mining method have been proposed. A high utility and interesting itemset is a subset of a high utility itemset, and its relations are strong. In this paper, we introduce the KWU-Mining algorithm to mine the high utility and interesting itemsets based on the WIT tree structure. The experiment results of the KWU-Mining algorithm show that they are effective in many transaction databases.