Supporting problem representations in decisions with strategic interactions

This paper argues that formalized procedures for decision making as they focus almost exclusively on the outcomes have limited utility. This is exposed particularly when the decision problem is complex and ill-structured and when there are strategic interactions between an agent and the environment. We propose an approach to support making this type of decision. This involves an agent structuring and restructuring the problem representation. The support system assists this effort to obtain a consistent and explainable interpretation of the problem and the environment. Rule-based formalism is proposed as a way of representing the agent's mental model of the problem, to modify this representation, and to integrate it with complementary models which traditionally have been used in decision support systems.