Conflict Resolution and Concensus Development among Inherently Contradictory Agents Using Fuzzy Linguistic Variables

Terrence P. Fries

Decision making based upon the recommendations of multiple intelligent agents has become common in various applications. However, difficulty arises when the agents have quite different recommendations. Many methods have been proposed that attempt to resolve conflicting opinions multiple, heterogeneous agents in decision making. However, all of these methods require that the agents negotiate until they arrive at a consensus opinion. These do not provide for the cases in which the agents have contradictory opinions that cannot be compromised. In certain cases, agent opinions will conflict due to the nature of the agents’ viewpoints. By forcing compromise or neglecting selected conflicting opinions, valuable information may be lost that adversely affect the decision. This paper proposes a method by which a consensus decision can be developed while not requiring that the individual agents abandon their opinions.

Full Text