CBR-based Explanation-aware Army Builder for Warhammer Fantasy Battle

摘要

Warhammer Fantasy Battle is a board game with two distinct phases, building army lists and using these armies in combat. The army list creation can be viewed as complex resource optimization problem, where the goal is to fully utilize the points available to build a good army. The goodnes of an army is often defined by how well the units of that army synergize with each other, and make up for drawbacks of other units. Thus some insight may be aquired from previous experiences in the form of well tested complete armies.To be able to propose complete armies to a user, and provide the opportunity to specify both units to be included and the number of points of this army, we have built a case-based reasoning (CBR) system. The system bases itself on reusing a previous army list and adapting this army until it meets the users requirements and stays within the game rules. This system was achieved by using the framework myCBR, its powerful modeling capabilities and retrieval engine. The results verify that the system is able to produce a varied set of valid solutions to the problem. Yet the goodness of these solutions have yet to be determined.