Abstract:
Association rule mining is one of the key issues in knowledge discovery. The discovery of frequent patterns, association, and correlation relationship among huge amounts of data is useful in selective marketing, decision analysis and business management. Association rules are traditionally defined as implications of the form A=>B, where A and B are frequent itemsets in a transaction database. The method extends traditional associations to include association rules of forms A => ¬B, ¬ A => B, and ¬ A => ¬ B, which indicate negative associations between itemsets. The negative rules are generated from infrequent itemsets. This system generates the set of frequent itemsets and the set of infrequent itemsets with three database including books, electronic and grocery store of sale transaction. This system presents a method for mining both positive and negative association rules. This system demonstrates that experimental results and efficiency of both positive and negative association rules.
