APPLICATION OF DATA ANALYSIS METHODS FOR AUTOMATION OF ONTOLOGY FORMATION

Abstract

Objectives. The aim of this work is to develop methods for automated text analysis and the retrieval of relevant data from full-text documents, as well as applying semantic text analysis methods for using linguistic ontologies as formalised models of subject area representation. Another aim is the use of electronic encyclopedias, primarily Wikipedia, as the basis for constructing the linguistic ontologies in order to derive maximum semantic information about their concepts, vocabulary expressions, interrelations and hierarchy.Methods.The search for solutions based on system analysis methods is based on the emergence of new technologies that for solving both the text itself and the object of research that is to be solved as a result of such processing. When creating contemporary artificial intelligence systems or their components, developers and researchers often face the need to formalise a certain subject area in order to automate the processing of phrases, word collocations and sentences entering the system in natural language form. Currently, the most popular approach to the formal description of a subject area is to construct an ontology.Results. Established approaches to the retrieval of information are described along with the architecture of the automated system and the results of their application.Conclusion. Semantic data analysis methods are applied with linguistic ontologies used as the formalised models of subject area representation. Approaches to retrieving information from Wikipedia are described along with the architecture of the automated system and results of its application.

Keywords

• automated system
• ontology
• subject area
• automated ontology construction
• wikipedia
• semantics