Title: An efficient approach to categorising association rules

Authors: Dongwoo Won; Dennis McLeod

Addresses: Semantic Information Research Laboratory, Computer Science Department, Viterbi School of Engineering, University of Southern California, Los Angeles, CA 90089-0781, USA. ' Semantic Information Research Laboratory, Computer Science Department, Viterbi School of Engineering, University of Southern California, Los Angeles, CA 90089-0781, USA

Abstract: Association rules are a fundamental data mining technique, used for various applications. In this paper, we present an efficient method to make use of association rules for discovering knowledge from transactional data. First, we approach this problem using an ontology. The hierarchical structure of an ontology defines the generalisation relationship for the concepts of different abstraction levels that are utilised to minimise the search space. Next, we have developed an efficient algorithm, hierarchical association rule categorisation (HARC), which use a novel metric called relevance for categorising association rules. As a result, users are now able to find the needed rules efficiently by searching the compact generalised rules first and then the specific rules that belong to them rather than scanning the entire list of rules.

Keywords: association rules; categorisation; data mining; market basket data; ontologies; relevance; transactional data.

DOI: 10.1504/IJDMMM.2012.049881

International Journal of Data Mining, Modelling and Management, 2012 Vol.4 No.4, pp.309 - 333

Received: 08 May 2021
Accepted: 12 May 2021

Published online: 18 Oct 2012 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article