An Agenda based Multi Issue Negotiation Model

Abstract

In multi-issue automated negotiation, group of agents may have different preferences and expectations on different issues. To reach an optimal negotiation outcome efficiently and effectively among a group of agents by considering their different preferences is a challenging problem in multi-issue automated negotiation. Negotiation convergence is a crucial part of the multi-issue automated negotiation process. To improve the negotiation convergence process and to minimise multi-issue negotiation complexity an agenda based approach is proposed to build a new negotiation model. The agenda based model investigates the significance of the choice of the sequence in issue-by-issue negotiation and also analyses the appropriateness of applying a preference ordering technique to generate the choice of sequence of issues. This model improves an agent’s utility by providing efficient convergence and an effective equilibrium outcome. The main motivation is to construct a new model which is symmetric and Paretooptimal. The proposed automated negotiation approach deals with multiple issues in such a way that it
helps to balance the efficiency and effectiveness of the outcome.

The negotiation is an interactive and decision making process by which a joint consensus decision is made by two or more parties. An agenda based negotiation approach is proposed where agents are led to cooperate in order to achieve a global goal while trying to satisfy as best as possible individual preferences and objectives. In this model multiple agents have their own decision measure to evaluate an offer related with both independent and sequentially interlinked multiple issues. This proposed approach shows how an issue-by-issue agenda based negotiation affects the negotiation outcome and makes it easy for the agents to find a common zone. The impact of the agenda procedure along with the strategy on the
outcome is illustrated by a case study. The agent’s preference order is used as a base case for providing the optimal agenda-procedure combination.

In a multi-issue negotiation process, it is difficult to reach an optimal outcome when the group of agents' preferences and the relative importance of the issues are not known to each other. This research investigates the multi-issue fuzzy group decision making problem where preference information on issues associated with services or products is expressed in different formats by the agents. A fuzzy preference concept is used to order the issues in the pre-negotiation stage to improve the efficiency of the outcome.
A dynamic and iterative group consensus process is developed to support the preference ordering process. Furthermore, the preferences of the decision-making agents are constructed using a multi-issuemethodology allowing the agents’ to take into account the preference ordering that can be used in the negotiation process, in order to help in quickening the search of a consensus between the agents. Therefore, the proposed negotiation approach consists of a multi-criteria decision making process and a
cooperation-based multi-agent multi-issue negotiation protocol. The proposed negotiation protocol and strategy reach a Pareto optimal agreement on multiple issues. Experimental results illustrate that the proposed model can help to make an agreement efficiently under diverse situations of conflict scenarios.

Date of Award	2014
Original language	English
Awarding Institution	Coventry University
Supervisor	Nick Godwin (Supervisor)