The Importance of Transformational Method in Automatic Idiom Translation
Abstract
This study aims to study the effectiveness of transformational methods in the process of automatic translation of idioms from the Uzbek language into English. The polysemy, cultural loading and context-dependent nature of phrases create significant problems in their machine translation. The study analyzed 3547 phraseological units taken from the Explanatory Dictionary of the Uzbek language and the uzschoolcorpora.uz national corpus, and their semantic equivalents in English were determined based on the British National Corpus (BNC) and the Oxford Dictionary of Idioms. The transformational model was developed based on syntactic structure, semantic similarity and statistical correspondence. The results show that the transformational approach has a significant advantage over literal translation in terms of BLEU (Bilingual Evaluation Understudy) and METEOR (Metric for Evaluation of Translation with Explicit Ordering) indicators, ensuring contextually and culturally accurate translation of idioms. This model can be implemented as an effective tool for improving the quality of automatic translation and enriching scientific approaches in the field of computational linguistics and translation studies.