Ontological Approach to Creating Viable Intelligent Systems

The modern challenge to artificial intelligence systems is to reduce the complexity of their development and maintenance during the life cycle, as well as the explanation of the solutions obtained. The paper shows that the ontological infrastructure that provides the creation and control of all components of systems with knowledge bases, can successfully ensure the fulfillment of these requirements. The paper describes approaches to solving the problem of controllability of this class of systems and, above all, the component most susceptible to changes - the knowledge base, describes an approach to the formation of ontological solvers from reusable components with the explanations generation. The proposed technology and its instrumental support are implemented on the IACPaaS platform, and a number of application systems for various domains have been developed using it.