Ontology Models in Intelligent System Engineering: A Case of the Knowledge-Intensive Application Domain

Abstract

The article describes the application of the ontological approach to intelligent system engineering. This approach suggests that the ontology models be presented in the form of interconnected modules of applied logic theories. This approach turns out to be effective in the case of a knowledge-intensive application domain, such as chemistry. Intelligent system that is being developed is supposed to solve the problem of organic compound reaction capacity identification. The problem is solved utilizing the concept systems of several chemistry subdomains. The ontology model is presented. The intelligent system model is provided. The analysis of the intelligent system requirements and interface quality attributes has brought into sharp focus several advantages of the utilized approach, i.e. the extensibility of the system due to the possibility to correct knowledge and metaknowledge during the system lifecycle, the potential to add problem solvers for new classes of tasks, and the increase in user confidence due to the utilization of user-understandable concept systems. These advantages become of paramount importance for the vitality of intelligent systems in the field where the intensification of knowledge-retrieval procedures and constant accumulation of knowledge (associated primarily with organic synthesis) make such knowledge more and more difficult for humans to conceive.