Implementing Association Rules Generator using CHARM

Abstract

In traditional association rule mining that based on Apriori-like algorithm, all frequent itemsets are computed and strong association rules are then generated from the set of all frequent itemsets as a next step. By contrast, in closed association rule mining, instead of mining all frequent itemsets in the first step, only the set of all closed frequent itemsets is mined where the size of the set of all closed frequent itemsets is much smaller than that of all frequent itemsets. Instead of generating rules from the set of all frequent itemsets, rules are generated from the set of all closed frequent itemsets , where the size of rules generated from the set of closed frequent itemsets is much smaller than that generated from the set of all closed frequent itemsets. This paper describes the concept of closed association rule mining, how CHARM algorithm works to find out all closed frequent itemsets, and the implementation of rules generating system or rules generator that mainly uses closed association rule mining techniques by using CHARM Algorithm.