Ijtimoiy-gumanitar fanlar

CHALLENGES IN RESEARCHING SYNTACTIC VALENCY IN NATURAL LANGUAGE PROCESSING

Computational linguistics, artificial intelligence, system and speech, valency theory, syntaxeme, component analysis, 7 syntactic relation operators, omoformia, transformation algorithm.

Authors

  • Rustam ASADOV PhD, dotsent, Samarqand akademik litseyi, Samarqand, O‘zbekiston, Uzbekistan

In this article, the problems of improving the accuracy of syntactic analysis in natural language processing (NLP) and artificial
intelligence systems are investigated from a philosophical-linguistic perspective. The linear limitations of Western computational
linguistics and the concept of "Language – Lison (System) – Norm – Speech" in Uzbek linguistics are compared, and the software
limitations of the binary opposition language/speech in English are proved. Based on the analysis of homonymic and contextual
errors resulting from the mixing of traditional levels, a system of 7 strict syntactic relation operators based on the theory of
professors A.M.Mukhin and U.U.Usmanov is proposed as a solution. Based on the selected empirical materials, the external
(speech-component) and internal (linguistic-syntaxemic) structures of the sentence are formally modeled.