THEORETICAL AND METHODOLOGICAL FOUNDATIONS OF DEVELOPING UZBEK–CHINESE PARALLEL CORPORA: LINGUISTIC ALIGNMENT, TRANSLATION EQUIVALENCE, AND CROSS-LINGUISTIC STRUCTURAL DIFFERENCES

The article examines the theoretical and methodological foundations of building Uzbek–Chinese parallel corpora. Although parallel corpora have become a core resource for translation studies, contrastive linguistics and machine translation, the Uzbek–Chinese language pair remains poorly resourced. The study analyzes three interrelated problems: the alignment of texts at the document, paragraph, sentence and word levels; the identification of translation equivalence; and the structural differences between an agglutinative language with subject–object–verb order and an isolating language with subject–verb–object order. Drawing on recent work on Uzbek corpus development and on embedding-based word alignment, the article proposes a multi-level framework for corpus construction that combines automatic alignment with linguistically informed annotation and human verification. The framework is intended to support contrastive research, translator training and the development of language technologies for the Uzbek–Chinese pair.