Past events ICDM´2002

IBaI Institute | Home | Past events | ICDM´2002 | Abstracts

Program past events

2nd Industrial Conference on Data Mining ICDM´2002
13. 06. - 15. 06. 2002 Leipzig/Germany

Principles of Association Rule Mining Algorithms

Dipl.-Inf. Jochen Hipp, Prof. Gholamreza Nakhaeizadeh, DaimlerChrysler AG, Research & Technology, Prof. Ulrich Güntzer, Wilhelm Schickard-Institute, University of Tübingen/Germany

Association rule mining is one of the fundamental data mining tasks. Since its introduction a broad variety of algorithms has been developed to cope with the computational very intensive generation of association rules.

Performance is crucial but unfortunately today the variety of available different association mining algorithms easily overwhelms even experts. The main problem is that typically the algorithms are all described separately without putting them into a general context.

In our paper we address this by a top-down approach. First of all we formalize and quantify the association rule problem in general. Then we work out the common principles of association rule mining. We do this independently of any particular approach. Based on these preliminaries we come down to the specifics of the actual algorithms.

The main contribution of this paper is an exhaustive systematization of today's established association rule mining algorithms. Moreover by for the very first time putting the algorithms into context differences and common aspects become obvious.

We complete our study by an evaluation of the algorithms on several different real world datasets and artifical test data. The insights presented in this paper lead to a deeper understanding of today's algorithms and their behavior on different datasets. Thus both practitioners and scientists can benefit from our work when deciding upon the application of association rule mining algorithms.



mission | ICDM´2009 | past events | publications | tutorial days | contact

Advances in Data Mining

Special Issue
Appeared: 2006

Advances in Data Mining

IBaI Publishing
ISSN: 1865-6781

Advances in Data Mining

Springer Verlag
ISBN: 978-3540707172

Data Mining and Multimedia Data

Springer Verlag
ISBN: 3-540-00317-7



Logfile-Analysetool NetLog

Logfile-Analysetool NetLog