| Abstract: | This study proposes a multi-agent-based model for bilateral multi-objective negotiation in electronic commerce trading. The function and behaviour of several types of agents are discussed. These agents interact with each other in our negotiation support system model to create the most appropriate solution for both negotiating parties. In particular, we are concerned with a win-win negotiation approach in which agents seek to strike a fair deal that also maximizes the payoff for everyone involved. That is, if the opponent cannot accept an offer then the proponent should endeavour to find an alternative to make a trade-off. Against this background, a utility model based on fuzzy constraint satisfaction problems is proposed to ensure that these agents reach a solution that is fair for both negotiating parties if such a solution exists. The model uses prioritized fuzzy constraints to indicate how concessions should be made when necessary. In addition, by incorporating the notion of a negotiation argument into our evaluation model, the agents can sometimes reach agreements that would otherwise be impossible. Finally, a numerical example is given to display the applicability of the proposed approach for electronic trading assistance. |
