Digital Library

cab1

 
Title:      IMPLICIT EXTRACTION OF USER CREDIBILITY FOR REPUTATION SYSTEMS IN E-COMMERCE
Author(s):      Jin Hyung Cho , Hwan Soo Kang , Kwiseok Kwon , Hwan Il Kang
ISBN:      972-8924-23-2
Editors:      Sandeep Krishnamurthy and Pedro Isaías
Year:      2006
Edition:      Single
Keywords:      reputation system, e-commerce, collaborative filtering, source credibility model, user credibility, implicit extraction
Type:      Full Paper
First Page:      183
Last Page:      190
Language:      English
Cover:      cover          
Full Contents:      click to dowload Download
Paper Abstract:      Online reputation systems are one of the most promising components for building trust among market participants in e-commerce. Conventional reputation systems rely on explicit rating feedback about users, which is evaluated by other users. However, in case of most e-commerce sites, users may refrain from providing sufficient explicit feedback or provide intentionally or unintentionally untruthful feedback. Therefore, if possible, it would be more practical to extract reputation value implicitly from users’ past ratings for transaction items or transaction data in order to overcome the limitations of explicit reputation mechanisms. We propose a novel reputation mechanism suitable for the e-commerce sites which have no sufficient explicit ratings and trust relationships among users. The conceptual framework is based on the source credibility model in consumer psychology. Collaborative filtering (CF) is adopted to quantify credibility factors for users’ reputation and predict general users’ opinions about transaction items. We experimentally evaluated the performance of the proposed system by comparing with other benchmark reputation mechanisms. The experimental results provide evidence that general users’ ratings can be predicted effectively by only a small number of evaluators using the proposed method.
   

Social Media Links

Search

Login