Title: Identifying buyers with similar seller rating models and using their opinions to choose sellers in electronic markets

Authors: Sandhya Beldona, Costas Tsatsoulis

Addresses: Department of Electrical Engineering and Computer Science, Information and Telecommunication Technology Center, The University of Kansas, 2335 Irving Hill Road, Lawrence, KS 66045, USA. ' Department of Computer Science and Engineering, University of North Texas, 1155 Union Circle # 310440, Denton, Texas 76203-5017, USA

Abstract: In this paper, we provide a model for designing buyers that can learn to identify trustworthy friends that are honest and share similar opinions in a decentralised electronic market. The buyer rates a seller after having purchased goods from it. It also evaluates friends who provide seller information when requested, to determine their truthfulness, similarity in opinions regarding product expectations, and to identify the differences between its own and its friends| seller rating mechanisms. Trustworthy friends are identified, and the seller ratings provided by them are adjusted to account for the differences in rating systems and then utilised to evaluate sellers. We conducted experiments to confirm that a buyer using the proposed model is able to accurately identify trustworthy friends, accurately adjust the seller reputation ratings provided by them and has higher gains than a buyer acting alone.

Keywords: autonomous agents; buyer agents; e-commerce; electronic commerce; electronic markets; opinion similarity; reputation; seller agents; seller decisions; seller information; seller rating models; trust; information science; decision science.

DOI: 10.1504/IJIDS.2010.029901

International Journal of Information and Decision Sciences, 2010 Vol.2 No.1, pp.1 - 16

Published online: 02 Dec 2009 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article