LEXICO-SEMANTIC AMBIGUITY IN THE ENGLISH-KARAKALPAK PARALLEL CORPUS AND ITS IMPACT ON THE QUALITY OF MACHINE TRANSLATION

Abstract

The article examines the problem of lexical-semantic ambiguity in parallel corpora and its impact on the quality of machine translation using the example of the English-Karakalpak language pair. Karakalpak is a low-resource Turkic language with agglutinative morphology, which increases the complexity of translation. It is shown that polysemy of lexical units is a key cause of semantic distortions, especially in conditions of limited linguistic resources. The necessity of integrating context-oriented and semantic methods into machine translation systems is substantiated.