NATIONAL-CULTURAL CHARACTERISTICS OF PEJORATIVE UNITS IN THE UZBEK AND RUSSIAN LANGUAGES AND THEIR TRANSLATION USING ARTIFICIAL INTELLIGENCE

Abstract

This article examines the national-cultural features of pejorative units in the Uzbek and Russian languages, as well as the problems of their translation using artificial intelligence technologies. The pragmatic, semantic, and linguocultural characteristics of negatively evaluative vocabulary are analyzed. Special attention is paid to the difficulties of conveying emotional coloring, irony, and cultural connotations in neural machine translation systems. Based on material from Uzbek and Russian, typical errors made by AI translators in interpreting pejorative expressions are identified. The research employs comparative, contextual, and linguopragmatic methods of analysis. As a result, the necessity of considering cultural context and pragmatic intention in the development of intelligent translation systems is substantiated.