Abstract

In the face-to-face bargaining process, people usually persuade opponents to accept offers by various strategies. It is a challenge to exhibit such persuasion capability in the on-line automated bargaining process for e-trade. In this paper, we propose a multi-agent architecture to perform on-line bargaining activities with persuasion capability. We demonstrate the usability of fuzzy technique to implement the on-line bargaining and persuasion system, called AutoPersuader. We also evaluated the effect of on-line persuasion and the influence of self-esteem by laboratory experiments. The experiments used the financial gain and customer satisfaction to evaluate AutoPersuader with and without persuasion capability. The results show that the quality of persuasive argumentation is a key factor that influences the usefulness of persuasion mechanism embedded in on-line bargaining system. The results also show that self-esteem may not be a critical factor in on-line persuasion situation.

Share

COinS