Web Usage Mining: Extraction, Maintenance and Behaviour Trends

With the growing popularity of the web, large volumes of data are gathered automatically by Web Servers and collected into access log files. Analysis of such files is generally called Web Usage Mining and tends to search user's behaviour patterns from one or more web servers in order to extract relationships within extracted data. Over the past few years, numerous work have focussed on this problem and some tools have been created to analyse user's behaviour on a web server. Even if it is possible to analyse user's behaviour, two important problems are not taken into account: how to handle new records and capitalize on previous knowledge and how to analyse trend user in behaviour? In this paper we present AUSMS-Web, a system which aims to extract knowledge from a user's behaviour, to maintain this knowledge when new data is added to the logs, and to analyse users’ behaviour trends on a web site.

• Type: Conference
• Year: 2003
• Where: IICAI
• Authors: Pierre-Alain Laur, Maguelonne Teisseire, Pascal Poncelet
• Keywords: IICAI 2003, IICAI 2007, User's Behaviour, User's Behaviour Patterns, Web Server