Profile reusability for personalized information access : application to users contexts determination.

Max Chevalier (1, 2) , Chantal Soulé-Dupuy (1, 3) , Pascaline Tchienehom (4)

1 IRIT-SIG - Systèmes d’Informations Généralisées
2 UT3 - Université Toulouse III - Paul Sabatier
3 UT Capitole - Université Toulouse Capitole
4 IRIT - Institut de recherche en informatique de Toulouse

The notion of context is important when we are dealing with personalization. Users often change their needs in time and the determination of contexts can help to identify those changes. In this article, we associate to each information or user a profile which describes it. We show that the descriptionof a profile can be done by re-using existing profiles and we propose a method to determine various types of users contexts based on structure, contents and re-usability of profiles. We discuss the interest of the suggested method through qualitative experiments carried out the collection "Los Angeles Times 94" of the CLEF evaluation campaign.

Mots clés: profile; granularity; Information Access; personalization; context

Domaines: Intelligence artificielle [cs.AI]; Informatique et langage [cs.CL]; Base de données [cs.DB]; Algorithme et structure de données [cs.DS]; Recherche d'information [cs.IR]; Langage de programmation [cs.PL]; Génie logiciel [cs.SE]; Web; Linguistique

Max Chevalier, Chantal Soulé-Dupuy, Pascaline Tchienehom. Profile reusability for personalized information access : application to users contexts determination.. International Workshop on Context-based Information Retrieval (CIR’05), in conjonction with Context’05, Jul 2005, Paris, France. p. 71-82. ⟨hal-01580415⟩