Issues in association rule mining and interestingness

Abstract

This work presents unaddressed issues in the field of Association Rule Mining (ARM).Looking at the previous literature of varied areas and applications of ARM, we identify three broad categories of ARM where the research is still in the nascent stage. We review papers in the three categories of fuzzy association rules, multilevel association rules and negative association rules to study the state-of-art research conceptually and algorithmically. As a result, we provide a compendium of gaps and unaddressed issues in these domains using our understanding of ARM and interestingness.