Clustering Users to Explain Recommender Systems' Performance Fluctuation

Recommender systems (RS) are designed to assist users by recommending them items they should appreciate. User based RS exploits users behavior to generate recommendations. Users act in accordance with different modes when using RS, so RS's performance fluctuates across users, depending on their act mode. Act here includes quantitative and qualitative features of user behavior. When RS is applied in an e-commerce dedicated social network, these features include but are not limited to: user's number of ratings, user's number of friends, the items he chooses to rate, the value of his ratings, and the reputation of his friends. This set of features can be considered as the user's profile. In this work, we cluster users according to their acting profiles, then we compare the performance of three different recommenders on each cluster, to explain RS's performance fluctuation across different users' acting modes.

[1]  J. Golbeck,et al.  FilmTrust: movie recommendations using trust in web-based social networks , 2006, CCNC 2006. 2006 3rd IEEE Consumer Communications and Networking Conference, 2006..

[2]  Robert Meersman,et al.  On the Move to Meaningful Internet Systems 2004: CoopIS, DOA, and ODBASE , 2004, Lecture Notes in Computer Science.

[3]  Donald W. Bouldin,et al.  A Cluster Separation Measure , 1979, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[4]  Anne Boyer,et al.  Hybridising Collaborative Filtering and Trust-aware Recommender Systems , 2012, WEBIST.

[5]  Bamshad Mobasher,et al.  Towards Trustworthy Recommender Systems : An Analysis of Attack Models and Algorithm Robustness , 2007 .

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

[7]  Harald Steck,et al.  Item popularity and recommendation accuracy , 2011, RecSys '11.

[8]  Gerald Quirchmayr,et al.  Multidisciplinary Research and Practice for Information Systems , 2012, Lecture Notes in Computer Science.

[9]  Jennifer Golbeck,et al.  Using probabilistic confidence models for trust inference in Web-based social networks , 2010, TOIT.

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

[11]  Georg Lausen,et al.  Spreading activation models for trust propagation , 2004, IEEE International Conference on e-Technology, e-Commerce and e-Service, 2004. EEE '04. 2004.

[12]  J. Golbeck.  Personalizing Applications through Integration of Inferred Trust Values in Semantic Web-Based Social Networks , 2005 .

[13]  Krista Rizman Zalik,et al.  An efficient k 0-means clustering algorithm , 2008 .

[14]  Anne Boyer,et al.  Analyzing Recommender System's Performance Fluctuations across Users , 2012, CD-ARES.

[15]  Robin Burke,et al.  Identifying Attack Models for Secure Recommendation , 2004 .

[16]  Bradley N. Miller,et al.  PocketLens: Toward a personal recommender system , 2004, TOIS.

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

[18]  Sanjay Ranka,et al.  An effic ient k-means clustering algorithm , 1997 .

[19]  Bobby Bhattacharjee,et al.  Using Trust in Recommender Systems: An Experimental Analysis , 2004, iTrust.

[20]  Alfred Kobsa,et al.  The Adaptive Web, Methods and Strategies of Web Personalization , 2007, The Adaptive Web.

[21]  Neil J. Hurley,et al.  Collaborative recommendation: A robustness analysis , 2004, TOIT.

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

[23]  Sophie Ahrens,et al.  Recommender Systems , 2012 .

[24]  John Riedl,et al.  Shilling recommender systems for fun and profit , 2004, WWW '04.

[25]  Kate Ehrlich,et al.  Pointing the way: active collaborative filtering , 1995, CHI '95.