Downloads
Abstract
In traditional mining of association rules, finding all association rules from databases that satisfy minSup and minConf faces with some problems in case of the number of frequent itemsets is large. Thus, it is necessary to have a suitable method for mining fewer rules but they still embrace all rules of traditional mining method. One of the approaches that is the mining method of essential rules: it only keeps the rule that its left hand side is minimal and its right side is maximal (follow in parent-child relationship). In this paper, we propose a new algorithm for mining the essential rules from the frequent closed itemsets lattice to reduce the time of mining rules. We use the parent-child relationship in lattice to reduce the cost of considering parent-child relationship and lead to reduce the time of mining rules.
Issue: Vol 12 No 11 (2009)
Page No.: 49-56
Published: Jun 15, 2009
Section: Article
DOI: https://doi.org/10.32508/stdj.v12i11.2311
Download PDF = 580 times
Total = 580 times