Title: A comparison of data preparation approaches for e-mail categorisation

Authors: Helmut Berger, Dieter Merkl, Michael Dittenbach

Addresses: E-Commerce Competence Center (EC3), Donau City Strasse 1, A-1220 Wien, Austria. ' Institut fur Softwaretechnik und Interaktive Systeme, Technische Universitat Wien, Favoritenstrasse 9-11/188, A-1040 Wien, Austria. ' E-Commerce Competence Center (EC3), Donau City Strasse 1, A-1220 Wien, Austria

Abstract: This paper reports on experiments in multi-class e-mail categorisation with supervised and unsupervised machine learning techniques. To this end, Support Vector Machines, decision tree learners, instance-based classifiers, Naive Bayes classification approaches and Self-Organising Maps were applied. A word-based and a character n-gram document representation approach were employed in order to assess the categorisation performance of the various learning approaches. The results indicate a substantial increase in classification accuracy when e-mail header information is considered in the document representation. To a much lesser degree, word-based document representations are advantageous over n-gram representations.

Keywords: e-mail categorisation; document representation; machine learning; indexing methods; information filtering; feature selection; data preparation; support vector machines; decision tree learners; instance-based classifiers; Bayes classification; self-organising maps; header information.

DOI: 10.1504/IJIIDS.2007.014946

International Journal of Intelligent Information and Database Systems, 2007 Vol.1 No.2, pp.91 - 121

Published online: 20 Aug 2007 *

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