Розроблення мовно–інформаційного інструментарію автоматизованого екстрагування з тексту семантичної й когнітивної інформації

The article deals with the role of semantic research for the creation of intelligent systems for automated processing of language information. An approach to the development of linguistic and informational tools for automated extraction of semantic and cognitive information from the text is outlined. The issue of increasing the efficiency of high-quality and fast processing of digital information, the creation of highly intelligent, fast-acting information systems was raised. The theoretical basis of the research is the lexicographic theory of symbolic semantic systems developed by academician V. A. Shyrokov and the theory of semantic states of natural language units built within it, which is based on the principles of lexical semantics and is oriented towards the use of computer technologies in semantic research. An integrated lexicographic system created in the Ukrainian Language and Information Fund of the National Academy of Sciences of Ukraine was selected as the source base of the study, the core component of which is the computer version of the «Dictionary of the Ukrainian Language» in 20 volumes, which is the most representative source of semantic information that provides a description of the lexical meaning of a word according to many parameters with the involvement of pragmatic, connotative, syntagmatic and linguistic contexts. In the study, the formalism of the theory of semantic states is used to build a linguistic and informational toolkit for automatic recognition of semantic and cognitive information in texts written in natural languages. The proposed semantic classifier provides an explicit presentation of the information extracted from the text, which enables its further software processing thanks to the unified interpretation of the operators of the semantic state formula and the results of the semantic indexing of the text. The expediency of using the developed language and information toolkit for solving such important tasks of computer linguistics as determining formal criteria for distinguishing lexical homonyms and LSV of a polysemous word, correct establishment of structural contrasts «meaning — shades», «lexeme — lexical-semantic variants» in interpreting the meaning of the latter, the task of unifying the description of the lexical meaning of words of one semantic field.