LINGUISTIC AND TECHNOLOGICAL FOUNDATIONS FOR BUILDING A LARGE-SCALE UZBEK TEXT CORPUS FOR NATURAL LANGUAGE PROCESSING
This article investigates the creation of a large-scale Uzbek text corpus for Natural Language Processing (NLP) systems within the implementation framework of Presidential Decree No. UP-6084 (October 20, 2020) and Presidential Resolutions No. PP-358 (October 14, 2024) and No. PP-308 (October 21, 2025). The study analyzes the impact of the agglutinative nature of the Uzbek language, complex graphemes of the current Latin alphabet (digraphs sh, ch, ng and apostrophe-modified letters o‘, g‘), and the parallel use of Cyrillic and Latin scripts on computational linguistics processes. The author proposes a corpus architecture encompassing text collection, orthographic normalization in Unicode (NFC), contextual transliteration, lemmatization, and morphological tagging. Based on the analysis of domestic and international computational linguistics research, actionable recommendations are formulated for developing a high-precision Uzbek text corpus.
Authors
- O'ralova Feruza O'ktam qizi
Publication Details
- Journal
- International Journal of Literature and Language Studies
- Published
- 2026-11-20
- Primary Topic
- Natural Language Processing Techniques
- Type
- article
- Field-Weighted Citation Impact
- 0.00