Source Credibility Model for Neighbor Selection in Collaborative Web Content Recommendation

Since collaborative filtering (CF) based recommendation methods rely on neighbors as information sources, their performance depends on the quality of neighbor selection process. However, conventional CF has a few fundamental limitations that make them unsuitable for Web content services: recommender reliability problem and no consideration of customers' heterogeneous susceptibility on information sources. To overcome these problems, we propose a new CF method based on the source credibility model in consumer psychology. The proposed method extracts each target customer's part-worth on source credibility attributes using conjoint analysis. The results of the experiment using the real Web usage data verified that the proposed method outperforms the conventional methods in the personalized web content recommendation.

[1]  Chunyan Miao,et al.  Improving collaborative filtering with trust-based metrics , 2006, SAC '06.

[2]  Paul Resnick,et al.  Reputation systems , 2000, CACM.

[3]  Audun Jøsang,et al.  A survey of trust and reputation systems for online service provision , 2007, Decis. Support Syst..

[4]  Aggelos Kiayias,et al.  Polynomial Reconstruction Based Cryptography , 2001, Selected Areas in Cryptography.

[5]  鄭宇庭.  行銷硏究 : Marketing research , 2009 .

[6]  Gediminas Adomavicius,et al.  Toward the next generation of recommender systems: a survey of the state-of-the-art and possible extensions , 2005, IEEE Transactions on Knowledge and Data Engineering.

[7]  Scott D. Johnson,et al.  Influences on consumer use of word-of-mouth recommendation sources , 1997 .

[8]  J. Carroll,et al.  Guest Editorial: Psychometric Methods in Marketing Research: Part I, Conjoint Analysis , 1995 .

[9]  H. R. Rao,et al.  A multidimensional trust formation model in B-to-C e-commerce: a conceptual framework and content analyses of academia/practitioner perspectives , 2005, Decision Support Systems.

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

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

[12]  Daniel W. Manchala.  E-Commerce Trust Metrics and Models , 2000, IEEE Internet Comput..

[13]  Rajiv Vaidyanathan,et al.  Eliciting Online Customers’ Preferences: Conjoint vs Self-Explicated Attribute-Level Measurements , 2003 .

[14]  Yongtae Park,et al.  Collaborative Filtering Using Dual Information Sources , 2007, IEEE Intelligent Systems.

[15]  Arvind K. Tripathi,et al.  Design of a shopbot and recommender system for bundle purchases , 2006, Decis. Support Syst..

[16]  Robert Wilensky,et al.  An algorithm for automated rating of reviewers , 2001, JCDL '01.

[17]  Barry Smyth,et al.  Trust in recommender systems , 2005, IUI.

[18]  Yoon Ho Cho,et al.  A personalized recommender system based on web usage mining and decision tree induction , 2002, Expert Syst. Appl..

[19]  P. Green,et al.  Thirty Years of Conjoint Analysis: Reflections and Prospects , 2001 .

[20]  Richard G. Netemeyer,et al.  Measurement of Consumer Susceptibility to Interpersonal Influence , 1989 .

[21]  M. Eisend.  Source Credibility Dimensions in Marketing Communication - A Generalized Solution , 2006 .

[22]  J. Douglas Carroll,et al.  Psychometric Methods in Marketing Research: Part I, Conjoint Analysis , 1995 .

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