Authors: Mohamed Khazri ¹; Mohamed Tmar ²; Mohamed Abid ¹; Mohand Boughanem ^3,4

• 1 Département de Génie Électrique de Sfax [ENIS]
• 2 Institut Supérieur d'Informatique et de Multimédia de Sfax
• 3 Systèmes d’Informations Généralisées
• 4 Université Toulouse III - Paul Sabatier

We propose in this paper an approach for document clustering. It consists of representing the corpus as a document graph, where the links are defined by some criteria. These links are quantified by simialrity measures. We aim join this context into the approach of classification to constitute small-worlds networks of homogeneous documents. The homogeneity of the clusters is measured according to the properties of small worlds. The clusters, as well as their proprietes, allow to rerank search results. Some experiments were done on a corpus provided by TREC and the obtained results show the contribution of small-worlds networks in information retrieval.