Search academic texts

Information × Registration Number 0219U101085, 0117U004327 , R & D reports Title Designing of a linguistic processor and a knowledge base as individual components of individual speech system to form a cluster of natural language information technology popup.stage_title Head Kyslenko Yuri I., Кандидат технічних наук Registration Date 08-05-2019 Organization National Technical University of Ukraine "Igor Sikorsky Kyiv Polytechnic Institute" popup.description2 The problem of developing a knowledge base and semantic Web, which Tim Berners-Lee spoke about twenty years ago, remains relevant, although no particular progress in its solution is evident. Since this problem is related to simulation of cognitive processes, available to us only through analysis of the speech level of their reproduction, the task of analyzing cognitive activity of a person at the linguistic level arises. In this work a method of automated processing of natural-language information is presented. The method is based on neurophysiological features of the speech device of a person, which allows to form a knowledge base on the basis of an arbitrary volume of natural-language texts. To this end, an integrational approach to the structural level of organization of language is used, which is based on the results of modern studies of speech activity of man in many areas of research. This approach eliminates many contradictions in classical grammar. The key idea of the ​​integrational approach is a concept of individual speech system, which consists of linguistic processor and knowledge base. The work describes functional load of the linguistic processor, including its specific procedures for text of mono- and poly-predicate levels. Also, the structure of knowledge of the natural-linguistic knowledge base is presented, which establishes formal connections at structural level between, on one hand, syntactic structure of a natural-language text and, on the other hand, semantic structure of its knowledge. Examples of use of the developed technology for synthesis of mono-predicate text and for natural-language search with a complex request are given. Product Description popup.authors popup.nrat_date 2020-04-02 Close