Abstract:

The article proposes the implementation of an automated system for ontologies comparing. The implementation of this system will simplify the process of document processing and determining their belonging to scientific fields. A possible implementation is considered, including a mechanism for automated ontology construction, which, based on the received document, will build an ontology, highlighting keywords and combining them by concepts. One of the advantages of using ontologies as the basis of the system is a systematic approach to the study of the subject area.

As an example, two ontologies on the subject of "Artificial Intelligence" were constructed, one built manually, the second generated automatically. In this case, it was important to consider possible differences in the construction of ontologies. Many ontologies are built by experts on the basis of their knowledge of the subject area and use the main works on a given topic. However, automatic construction of ontologies using statistical or weight characteristics of the text may differ significantly from the results provided by a person. That is why there is interest in how manual and automatic ontologies can be compared. First of all, the comparison of ontologies was carried out visually by an expert in order to highlight the main features of human information processing. As the main methods for comparing ontologies, the use of comparison of graphs and sets is considered.

Since the semantic network can be considered as an element of artificial intelligence systems. Then the proposed system for comparing ontologies can be taken into account as an intelligent decision support system or as its element. Intelligent decision support systems are an automated computer network that provides objective data analytics with the construction of a mathematical model of the expected development of events. The purpose of such products is to help people making complex management decisions in difficult conditions. The structure of an intelligent system includes three main blocks - a knowledge base, a decision inference mechanism, and an intelligent interface. In the future, such a system can be used in the education system. For example, to control the quality and level of education, or as part of a plagiarism check system.

Acknowledgments:
This paper is a part of the research work carried out within the Bauman Deep Analytics project of the Priority 2030 program.

Keywords:

Semantic networks, Ontologies, Artificial Intelligence, Big data.