Hierarchical grouping of association rules and its application to a real-world domain

Abstract

One common problem in association rule mining is that often a very large number of rules are generated from the database. The sheer volume of these rules makes it difficult, if not impossible, for human users to analyze and make use of the rules. In this article, we propose two algorithms for grouping and summarizing association rules. The first algorithm recursively groups rules according to the structure of the rules and generates a tree of clusters as a result. The second algorithm groups the rules according to the semantic distance between the rules by making use of a semantic tree-structured network of items. We propose an algorithm for automatically tagging the semantic network so that the rules can be represented as directed line segments in a two-dimensional space and can then be grouped according to the distance between line segments. We also present an application of the two algorithms, in which the proposed algorithms are evaluated. The results show that our grouping methods are effective and produce good grouping results.