Title: Efficient modelling of individual consumer preferences: facilitating agent-based online markets

Authors: Frank A. LoPinto, Cliff T. Ragsdale

Addresses: Management Department, La Salle University, 1900 West Olney Ave., Philadelphia, PA 19141, USA. ' Department of Business Information Technology, Virginia Polytechnic Institute and State University, 1007 Pamplin Hall, Blacksburg, VA 24061, USA

Abstract: As business-to-consumer e-commerce matures, better tools are needed to help consumers locate products they desire. Currently, user-driven searching, sorting and filtering are widely employed to align consumers and products. These processes could be improved by a quick and easy way to capture and model individual consumer preferences for particular product attributes. We present such a methodology using conjoint analysis and neural networks. This paper presents a study where preferences are elicited using a survey instrument, models of individual participants| preferences are constructed, and prediction accuracy is reported. This proposed methodology performs well for individuals in the example explored here.

Keywords: efficient modelling; individual preferences; consumer preferences; preference modelling; conjoint analysis; agent-based systems; multi-agent systems; online markets; neural networks; ANNs; business-to-consumer; B2C; e-commerce; electronic commerce; product attributes.

DOI: 10.1504/IJEMR.2010.030508

International Journal of Electronic Marketing and Retailing, 2010 Vol.3 No.1, pp.66 - 81

Published online: 20 Dec 2009 *

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