An empirical study of qualities of association rules from a statistical view point

Abstract

Minimum support and confidence have been used as the criteria for generating association rules in all association rule mining algorithms. These criteria have their natural appeals, such as simplicity, and few researchers have suspected the quality of generated rules. In this paper, we examine the rules from a more rigorous point of view by conducting statistical testing. Specifically, we use contingency tables and Chi-square test to analyze the data. The experimental results showed that one third of association rules based on the support and confidence criteria were not significant, that is, the antecedent and the consequent of the rules were not correlated. It indicates that minimum support and minimum confidence do not provide adequate discovery of useful associations. Chi-square statistics can be considered as an enhancement or an alternative solution.