Finding Representative Association Rules from Large Rule Collections.

Abstract One of the most well-studied problems in data mining is com- puting association rules from large transactional databases. Often, the rule collections extracted from existing data- mining methods can be far too large to be carefully examined and understood by the data analysts. In this paper, we address exactly this issue of over- whelmingly large rule collections by introducing and study- ing the following problem: Given a large collection R of association rules we want to pick a subset of them S µ R that best represents the original collection R as well as the dataset from which R was extracted. We flrst quantify the notion of the goodness of a ruleset using two very simple and intuitive deflnitions. Based on these deflnitions we then formally deflne and study the corresponding optimization problems of picking the best ruleset S µ R. We propose al- gorithms for solving these problems and present experiments to show that our algorithms work well for real datasets and lead to large reduction in the size of the original rule collec- tion.