Using self-defined group activities for improvingrecommendations in collaborative tagging systems

This paper aims to combine information about users' self-defined social connections with traditional collaborative filtering (CF) to improve recommendation quality. Specifically, in the following, the users' social connections in consideration were groups. Unlike other studies which utilized groups inferred by data mining technologies, we used the information about the groups in which each user explicitly participated. The group activities are centered on common interests. People join a group to share and acquire information about a topic as a form of community of interest or practice. The information of this group activity may be a good source of information for the members. We tested whether adding the information from the users' own groups or group members to the traditional CF-based recommendations can improve the recommendation quality or not. The information about groups was combined with CF using a mixed hybridization strategy. We evaluated our approach in two ways, using the Citeulike data set and a real user study.

[1]  Antal van den Bosch,et al.  Recommending scientific articles using citeulike , 2008, RecSys '08.

[2]  Sung-Bae Cho,et al.  Restaurant Recommendation for Group of People in Mobile Environments Using Probabilistic Multi-criteria Decision Making , 2008, APCHI.

[3]  Jun Ozawa,et al.  Cellular phone ringing tone recommendation system based on collaborative filtering method , 2003, Proceedings 2003 IEEE International Symposium on Computational Intelligence in Robotics and Automation. Computational Intelligence in Robotics and Automation for the New Millennium (Cat. No.03EX694).

[4]  Liliana Ardissono,et al.  Intrigue: Personalized recommendation of tourist attractions for desktop and hand held devices , 2003, Appl. Artif. Intell..

[5]  John G. Breslin,et al.  The Future of Social Networks on the Internet: The Need for Semantics , 2007, IEEE Internet Computing.

[6]  John Riedl,et al.  PolyLens: A recommender system for groups of user , 2001, ECSCW.

[7]  Peter Brusilovsky,et al.  Interest Similarity of Group and the Members: the Case study of Citeulike , 2010 .

[8]  Wu-Jun Li,et al.  TagiCoFi: tag informed collaborative filtering , 2009, RecSys '09.

[9]  Paolo Avesani,et al.  Trust-Aware Collaborative Filtering for Recommender Systems , 2004, CoopIS/DOA/ODBASE.

[10]  Ido Guy,et al.  Personalized recommendation of social software items based on social relations , 2009, RecSys '09.

[11]  Dan Frankowski,et al.  Collaborative Filtering Recommender Systems , 2007, The Adaptive Web.

[12]  J. Guare,et al.  The Future of Social Networks on the Internet , 2007 .

[13]  Shengchao Ding,et al.  Augmenting collaborative recommender by fusing explicit social relationships , 2009 .

[14]  Robin D. Burke,et al.  Hybrid Web Recommender Systems , 2007, The Adaptive Web.

[15]  Peter Brusilovsky,et al.  Social networks and interest similarity: the case of CiteULike , 2010, HT '10.

[16]  John Riedl,et al.  Application of Dimensionality Reduction in Recommender Systems , 2000 .

[17]  Kristian J. Hammond,et al.  Flytrap: intelligent group music recommendation , 2002, IUI '02.

[18]  Judith Masthoff,et al.  Effective explanations of recommendations: user-centered design , 2007, RecSys '07.

[19]  Rashmi R. Sinha,et al.  Comparing Recommendations Made by Online Systems and Friends , 2001, DELOS.