Research on mining positive and negative association rules

Positive and negative association rules are important to find useful information hided in massive data sets, especially negative association rules can reflect mutually exclusive correlaiton among items. Despite a great deal of research, a number of challenges still exist in mining positive and negative association rules. In order to solve the problem of “difficult to determine frequent itemsets” and “how to delete contradictive positive and negative association rules”, the paper presents a new algorithm for mining positive and negative association rules. The algorithm applies a new measurement framework of support and confidence to solve the problems existing. The performance study shows that the method is highly efficient and accurate in comparison with other reported mining methods.