THEORETICAL AND PRACTICAL FOUNDATIONS OF CREATING A LEGAL QUESTION-AND-ANSWER SYSTEM IN THE UZBEK LANGUAGE BASED ON THE EXPERIENCE OF TURKIC LANGUAGES
The article analyzes NLP resources (BERTurk, KazNERD, UD Treebanks) formed in the Turkish and Kazakh languages and
proposes a theoretical and practical model for creating a legal question-and-answer (Legal QA, LQA) system in the Uzbek language
based on them. The architecture is based on a hybrid principle: IR Retrieval (BM25/dense), LLM Reasoning (transformers) and
Rule-based Constraint Filtering (normative checking). The agglutinative morphology of the Uzbek language, SOV order, formulaic
constructions of official-legal discourse and the integration of annotation requirements based on UD into LQA are covered. As a
result, an integrated architecture, corpus and annotation requirements roadmap for Uzbek LQA was developed.
1. O‘zbekiston Respublikasi Vazirlar Mahkamasi. Raqamli iqtisodiyotni rivojlantirish bo‘yicha qo‘shimcha chora-tadbirlar to‘g‘risida: PQ-6079-son
qaror, 2020-yil 5-oktabr. – Elektron ma’lumot. – URL: https://lex.uz/docs/5030048 (murojaat qilingan sana: 11.12.2025).
2. GitHub - turkish-bert by stefan-it. – Elektron resurs. – URL: https://github.com/stefan-it/turkish-bert (murojaat qilingan: 11.12.2025).
3. GitHub - KazNERD by IS2AI. – Elektron resurs. – URL: https://github.com/IS2AI/KazNERD (murojaat qilingan: 11.12.2025).
4. Universal Dependencies - Kazakh. – Elektron resurs. – URL: https://universaldependencies.org/kk/ (murojaat qilingan: 11.12.2025).
5. Universal Dependencies - Uzbek. – Elektron resurs. – URL: https://universaldependencies.org/uz/index.html (murojaat qilingan: 11.12.2025).
6. Çöltekin, Ç. A Corpus of Turkish Offensive Language on Social Media // Proceedings of the Twelfth Language Resources and Evaluation Conference.
– Marseille, France, 2020. – P. 6174–6184. – European Language Resources Association.
7. Rakhimova, D.; Turarbek, A.; Karyukin, V.; Sarsenbayeva, A.; Alieyev, R. Legal AI in Low-Resource Languages: Building and Evaluating QA
Systems for the Kazakh Legislation // Computers. – 2025. – Vol. 14. – Art. 354. – DOI: https://doi.org/10.3390/computers14090354.
8. kurs ishi. - Elektron manba. - URL: https://www.scribd.com/document/920374070/Grammatik-Kategoriyalar-Va-Ularning-Uslubiy-Xususiyatlari
(murojaat qilingan sana: 11.12.2025).
9. Lex.uz: rasmiy huquqiy ma’lumotlar portali. - Elektron manba. - URL: https://lex.uz/ (murojaat qilingan sana: 11.12.2025).
Copyright (c) 2025 «ACTA NUUz»
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
.jpg)
1.png)
