A Model of Deceit-Tolerant Automated Negotiation for Open Environment

Agent-based automated negotiation has been widely advocated to substitute human negotiation in the areas of electronic commerce and cooperative computing. However, deceptive behaviors in negotiation process may cause great harm to users' interests and hinder the practical application of automated negotiation. This paper presents a deceit-tolerant automated negotiation (DTAN) model, which copes with the deception problems in a tolerant manner to make automated negotiation more adaptive and robust in open environment. Inspired by the intelligent processing in the human negotiation, the DTAN model functions with cooperative works of four constituents: active deceit detection (ADD); deceit-sensed decision (DSD); deceit precontrol (DP); and subliminal self-protection (SSP). The frameworks and principles of the DTAN model are presented in this paper