Causal mental model of decision in the task of building explanations in intelligent information system

The subject of research is the process of building causal mental models of intelligent system decisions. The aim is to develop a causal approach to building mental models of decisions in artificial intelligence systems to enable the construction and refinement of explanations according to the user's knowledge of the domain. Tasks: structuring mental models considering their properties in the aspect of forming explanations in intelligent systems; development of a causal mental model of intelligent system decision. Conclusions. The analysis of possibilities for using mental models in the task of building explanations has been performed, highlighting causal, analogical, folk, and dynamic models. The scientific novelty of the obtained results is as follows. A causal mental model of intelligent system decision is proposed, which contains a set of alternative representations of causes and effects of the obtained decision, taking into account the possibilities and limitations regarding the use of the decision. The model takes into account both positive and negative properties of the obtained result, which creates conditions for increasing the level of user trust. A formal representation of the causal mental model has been developed, containing temporally ordered rules that link input data as causes and properties of the obtained decision as effects. The rules are combined into alternatives, which makes it possible to select explanations according to input data constraints and limitations on the possibilities of using the intelligent system decision.