The development of an intelligent decision support system in a scaled-down simulation environment: A Conflict Resolution Intelligent System for the Informed Strategy Expert (CRISIS-Expert)

Resource management and conflict resolution skills are essential to navigating decisions in life, but many of us acquire them only in a piecemeal way and indirectly over a lifetime of decision-making. In terms of artificial intelligence (AI), a conflict resolution system is an important part of many intelligent systems. The resolution of problems in this discipline, with respect to almost all types of decision-making processes by the use of AI methods, presents a great challenge. There has been significant research on AI-related games during the last decade. Currently, intelligent elements found in game AI are provided as optimal advice, best practice solutions, or recommendations by a human counterpart. However, human expertise is too complicated to duplicate, and the skill sets associated with it are not easily transposed from human format to machine format. This research explores the design, technical aspects, and outcome of an intelligent decision support systems (IDSS) shell. From this, a proof-of-concept prototype is developed in order to demonstrate the viability of providing decision support in the form of expert advice over a range of strategic recommendations and actions. The primary aim is to help resolve problems, specifically, the resolution of conflicts involving resource management. In this project, a serious war game, Memoir ’44, is used as the scaled-down model. The performance and results from the different core inference engines are compared to the historical outcome of actual military conflicts. The outcome of the test results is then presented as data visualisation elements to facilitate understanding and comprehension by the human participant. The Conflict Resolution Intelligent System for the Informed Strategy Expert (CRISIS-Expert) is the final deliverable output of the proposed project. The CRISIS-Expert incorporates innovative and complementary computational intelligence methods that provide core software technologies which support the functionality of the proposed IDSS shell. The system is trialled (viii) in a scale-down problem arena. This approach has been adopted because the types of problems that involve conflict resolution are complex and too large to be resolved in a full-size operation within the scope of this project. Therefore, the problem space has been reduced into a flexible framework by scaling it down into a model with only the critical parameters representing the real-world problem. Such models exist in the form of serious games. Serious games are gamified versions and accurate models of their more extensive and more complex counterparts. The approach taken to decision-making in solving problems in the game environment is in many ways similar to solving the problems in the real world counterpart. Therefore, once the solutions to the scaled-down approach can be determined or optimised, the methods and processes can be upscaled to their corresponding larger components.