A decision support system for conflict resolution

Abstract

The major objective of this thesis is to develop a comprehensive decision support system (DSS) for systematically studying real-world strategic conflicts. The main components include the design and implementation of flexible interfaces for formulation and intepretation; construction of novel algorithms for modeling and analyzing small, medium and large-scale conflicts; development of new theoretical ideas in conflict resolution such as an innovation approach to coalition analysis; and incorporation of these new concepts in the DSS. The DSS is based upon existing and new research developments for the Graph Model for Conflict Resolution, and is referred to as GMCR II.

The option form is improved and extensively utilized to represent a conflict in terms of decision makers, options and preferences. Specially designed data structures and corresponding algorithms are implemented for generating possible states, removing infeasible states, coalescing indistinguishable states, and modeling allowable state transitions. For small disputes, a graph-based approach to modeling a conflict is suggested as an input format. Algorithms based on different approaches to facilitate preference elicitation are designed and implemented. Moreover, the DSS permits the choices of different decision makers in a model to be analyzed using different behavior patterns or solution concepts. A range of useful follow-up analyses allow many "what if" questions to be investigated and thereby provides an enhanced understanding of the conflict under study, which should in turn lead to better decisions.