[CBL05a] Web Usage Mining : WWW Pages classification from Log files

Conférence Internationale avec comité de lecture : International Conference on Machine Intelligence ACIDCA-ICMI'2005 , January 2005, Tozeur, Tunisia,

Auteurs: M. Charrad , M. Ben Ahmed , Y. Lechevallier

Mots clés: Web Usage Mining, Clustering, Web Mining, Data Mining, Machine Learning

Résumé: With the growth of web-based applications, there is significant interest in analyzing web usage data to better understand web usage, and apply the knowledge to better serve users. An important component of web personalization is the extraction of frequent paths and browsing patterns of a web user from web serverÂ’s log. The approach proposed in this paper consists of three steps: Preprocessing, construction of web site topology from web logs and web pages classification. In Preprocessing, pages on web site are processed to be organized into sessions which represent units of interaction between web users and the web server. The construction of web site topology consists in creating in internal representation of the site from logs to extract frequent paths. In page classification, parameters are introduced from pages access information and a principal component analysis is applied to differentiate content pages from auxiliary ones. The experiments on a real data set show that the approach is effective for web pages clustering.

Equipe: msdma

BibTeX

@inproceedings {
	CBL05a,
	title	=	"{Web Usage Mining : WWW Pages classification from Log files}",
	author	=	" M. Charrad and M. Ben Ahmed and Y. Lechevallier ",
	booktitle	=	"{International Conference on Machine Intelligence ACIDCA-ICMI'2005 }",
	year	=	2005,
	month	=	"January",
	address	=	"Tozeur, Tunisia",
}