An Efficient Method for Mining Frequent Weighted Closed Itemsets from Weighted Item Transaction Databases

1 Division of Data Science, Ton Duc Thang University, Ho Chi Minh, Viet Nam 4 2 Faculty of Information Technology, Ton Duc Thang University, Ho Chi Minh, Viet Nam 5 [email protected], [email protected] 6 7 Abstract: In this paper, a method for mining frequent weighed closed itemsets (FWCIs) 8 from weighted item transaction databases is proposed. The motivation for FWCIs is that 9 frequent weighted itemset mining, as frequent itemset (FI) mining, typically results in a 10 substantial number of rules, which hinders simple interpretation or comprehension. 11 Furthermore, in many applications, the generated rule set often contains many redundant 12 rules. The inspiration for FWCIs is that one potential solution to the rule interpretation 13 problem is to adopt frequent closed itemset. This study first proposes two theorems and a 14 corollary. One theorem is used for checking non-closed itemsets while joining two 15 itemsets to create a new itemset and the other theorem is used for checking whether a new 16 itemset is non-closed itemset or not. The corollary is used for checking non-closed 17 itemsets when using Diffsets. Based on these theorems and corollary, an algorithm for 18 mining FWCIs is proposed. Finally, a Diffset-based strategy for the efficient computation 19 of the weighted supports of itemsets is described. A complete evaluation of the proposed 20 algorithm is presented. 21