Combining social and semantic information for recommandation : comparative study

In this paper we present algorithms for recommender systems. Our algorithms rely on a semantic relevance measure and social network centrality measures to partially explore the network using depth-first search and breath-first search strategies. We implement and compare several social network centrality measures. We apply our algorithms on real dataset: the MovieLens one. Our results show that having algorithms combining degree and betweenness give high precision and recall values. Moreover, the importance of our algorithms rely on the fact that these algorithms explore a small part of the graph instead of exploring all the graph as the classical searching methods do. RÉSUMÉ. Dans cet article nous présentons des algorithmes pour la recommandation dans les réseaux sociaux. Ces algorithmes combinent, les mesures centralités dans les réseaux sociaux et les profils sémantiques des utilisateurs dans le processus de l’élaboration de la recommandation. Nous intégrons des heuristiques dans l’exploration de graphe (parcours en profondeur DFS et parcours on largeur BFS). Nous avons appliqué ces algorithmes sur un ensemble de données réelles extraits des données de MovieLens. Nos résultats montrent des valeurs de précision, de rappel et de F-measure satisfaisantes.

[1]  Pattie Maes,et al.  Social information filtering: algorithms for automating “word of mouth” , 1995, CHI '95.

[2]  Stephen P. Borgatti,et al.  Centrality and network flow , 2005, Soc. Networks.

[3]  Ronald L. Rivest,et al.  Introduction to Algorithms, third edition , 2009 .

[4]  Alessandro Micarelli,et al.  User Profiles for Personalized Information Access , 2007, The Adaptive Web.

[5]  John Riedl,et al.  GroupLens: an open architecture for collaborative filtering of netnews , 1994, CSCW '94.

[6]  Martin Ester,et al.  TrustWalker: a random walk model for combining trust-based and item-based recommendation , 2009, KDD.

[7]  Sihem Amer Yahia,et al.  A Study of the Benet of Leveraging Tagging Behavior to Model Users' Interests in del.icio.us , 2007 .

[8]  Naren Ramakrishnan,et al.  Studying Recommendation Algorithms by Graph Analysis , 2003, Journal of Intelligent Information Systems.

[9]  Bamshad Mobasher,et al.  Personalized recommendation in social tagging systems using hierarchical clustering , 2008, RecSys '08.

[10]  Robert E. Tarjan,et al.  Depth-First Search and Linear Graph Algorithms , 1972, SIAM J. Comput..

[11]  Philip Resnik,et al.  Using Information Content to Evaluate Semantic Similarity in a Taxonomy , 1995, IJCAI.

[12]  Philip S. Yu,et al.  Horting hatches an egg: a new graph-theoretic approach to collaborative filtering , 1999, KDD '99.

[13]  Jonathan L. Herlocker,et al.  Evaluating collaborative filtering recommender systems , 2004, TOIS.

[14]  Mark Newman,et al.  Networks: An Introduction , 2010 .

[15]  Martin G. Everett,et al.  A Graph-theoretic perspective on centrality , 2006, Soc. Networks.

[16]  Lars Schmidt-Thieme,et al.  Taxonomy-driven computation of product recommendations , 2004, CIKM '04.

[17]  Dominique Laurent,et al.  Exploiting Semantic and Social Information in Recommendation Algorithms , 2012, ISIP.

[18]  L. Freeman.  Centrality in social networks conceptual clarification , 1978 .