User Interest Extraction based on Weighted Tags

Collaborative tagging systems are based on assigning keywords freely chosen by users, which promotes ressources sharing and organization by the way and improves the information retrieval. The tags allocation by users is illustrated particularly in sites sharing photos or videos (Flickr, YouTube). As navigations and clicks, tags can be good indicators of the user's interests. In this paper, we examine the limitations of previous tag-based profile extraction. We believe that for a better result, tags of a resource must represent well its content. Existing systems consider ‘Popularity’ as the unique criterion to judge the tag effectiveness. But it does not always reflect its importance and representativeness to the resource. In this paper, we propose a novel approach based on tag strength to represent a user. In which we introduce weighted tags based on user expertise.

[1]  Rui Li,et al.  Survey on social tagging techniques , 2010, SKDD.

[2]  Steve Cayzer,et al.  Adaptive User Profiles , 2009 .

[3]  Jane Yung-jen Hsu,et al.  You Are What You Tag , 2008, AAAI Spring Symposium: Social Information Processing.

[4]  Christoph Trattner,et al.  On the Navigability of Social Tagging Systems , 2010, 2010 IEEE Second International Conference on Social Computing.

[5]  Vittorio Loreto,et al.  Network properties of folksonomies , 2007, AI Commun..

[6]  Adam Mathes,et al.  Folksonomies-Cooperative Classification and Communication Through Shared Metadata , 2004 .

[7]  Federica Cena,et al.  Towards a Tag-Based User Model: How Can User Model Benefit from Tags? , 2007, User Modeling.

[8]  Steve Cayzer,et al.  Learning User Profiles from Tagging Data and Leveraging them for Personal(ized) Information Access , 2007, WWW 2007.

[9]  Qi Gao,et al.  Analyzing Cross-System User Modeling on the Social Web , 2011, ICWE.

[10]  Michael Granitzer,et al.  Learning user characteristics from social tagging behavior , 2012, HT '12.

[11]  Laks V. S. Lakshmanan,et al.  Efficient network aware search in collaborative tagging sites , 2008, Proc. VLDB Endow..

[12]  Hakima Mellah,et al.  Tags Weighting Based on User Profile , 2011, AMT.

[13]  Federica Cena,et al.  Tag-based user modeling for social multi-device adaptive guides , 2008, User Modeling and User-Adapted Interaction.

[14]  Roelof van Zwol,et al.  Flickr tag recommendation based on collective knowledge , 2008, WWW.

[15]  Bernardo A. Huberman,et al.  The Structure of Collaborative Tagging Systems , 2005, ArXiv.

[16]  Umberto Straccia,et al.  User Profile Modeling and Applications to Digital Libraries , 1999, ECDL.

[17]  Mokrane Bouzeghoub,et al.  Personnalisation de l'information: aperçu de l'état de l'art et définition d'un modèle flexible de profils , 2005, CORIA.

[18]  Lynda Tamine,et al.  Approche statistique pour la définition du profil d'un utilisateur de système de recherche d'information , 2007 .

[19]  Giovanni Quattrone,et al.  A query expansion and user profile enrichment approach to improve the performance of recommender systems operating on a folksonomy , 2010, User Modeling and User-Adapted Interaction.

[20]  Wolfgang Nejdl,et al.  The Benefit of Using Tag-Based Profiles , 2007, 2007 Latin American Web Conference (LA-WEB 2007).

[21]  Corinne Amel Zayani.  Contribution à la définition et à la mise en oeuvre de mécanismes d'adaptation de documents semi-structurés , 2008 .