Semantic Analysis of Uzbek-Language Texts Using Natural Language Processing: A Methodological Framework and Evaluation Protocol
Keywords:
natural language processing, Uzbek language, semantic analysis, agglutinative morphology, tokenization, word embeddings, BERT, text classificationAbstract
Uzbek is a low-resource, highly agglutinative Turkic language written in both Latin and Cyrillic scripts. Its productive suffixation and context-dependent word senses make automatic semantic analysis harder than for analytic languages such as English.
This study designs and examines a complete natural language processing (NLP) pipeline for the semantic analysis and topical classification of Uzbek texts, and asks which pipeline components matter most for meaning-sensitive tasks.
We specify a five-stage pipeline: normalization and script handling, tokenization, morphological analysis, vector representation (Word2Vec, fastText, BERT-type encoders) and supervised classification. We combine a worked linguistic analysis of representative Uzbek sentences with a structured synthesis of peer-reviewed and preprint benchmarks for Uzbek NLP, and we define a reproducible evaluation protocol.
Morphological decomposition reduces inflected surface forms to shared stems, for example talaba, talabalar and talabalarimizdan. Contextual encoders separate senses of homonyms such as yuz. Published benchmarks show that monolingual Uzbek BERT models outperform multilingual BERT in topic classification (F1 85.2 versus 83.4) and part-of-speech tagging (accuracy 91.6% versus 80.2% to 86.0%), and that rule-based morphological analysis reaches over 91% word-level accuracy.
Reliable Uzbek semantic analysis depends on morphology-aware preprocessing together with contextual embeddings trained on clean, script-consistent corpora. No new experimental measurements are reported here; the paper supplies a framework and protocol for them
References
1. Mansurov, B., & Mansurov, A. (2021). UzBERT: pretraining a BERT model for Uzbek. arXiv:2108.09814.
2. Mansurov, B., & Mansurov, A. (2020). Development of word embeddings for Uzbek language. arXiv:2009.14384.
3. Kuriyozov, E., Vilares, D., & Gómez-Rodríguez, C. (2024). BERTbek: A pretrained language model for Uzbek. In Proceedings of the 3rd Annual Meeting of the Special Interest Group on Under-resourced Languages (SIGUL 2024) (pp. 33–44). ELRA.
4. Kuriyozov, E., Salaev, U., Matlatipov, S., & Matlatipov, G. (2023). Text classification dataset and analysis for Uzbek language. In Proceedings of the 10th Language & Technology Conference, Poznań. arXiv:2302.14494.
5. Bobojonova, L., Akhundjanova, A., Ostheimer, P. S., & Fellenz, S. (2025). BBPOS: BERT-based part-of-speech tagging for Uzbek. In Proceedings of LoResLM 2025. ACL Anthology 2025.loreslm-1.23.
6. Salaev, U. (2024). UzMorphAnalyser: A morphological analysis model for the Uzbek language using inflectional endings. AIP Conference Proceedings, 3244(1), 030058. https://doi.org/10.1063/5.0241461
7. Mansurov, B., & Mansurov, A. (2021). Uzbek Cyrillic-Latin-Cyrillic machine transliteration. arXiv preprint.
8. Mikolov, T., Chen, K., Corrado, G., & Dean, J. (2013). Efficient estimation of word representations in vector space. arXiv:1301.3781.
9. Bojanowski, P., Grave, E., Joulin, A., & Mikolov, T. (2017). Enriching word vectors with subword information. Transactions of the Association for Computational Linguistics, 5, 135–146.
10. Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2019). BERT: Pre-training of deep bidirectional transformers for language understanding. In Proceedings of NAACL-HLT 2019 (pp. 4171–4186).
11. Conneau, A., et al. (2020). Unsupervised cross-lingual representation learning at scale. In Proceedings of ACL 2020 (pp. 8440–8451).
12. Sennrich, R., Haddow, B., & Birch, A. (2016). Neural machine translation of rare words with subword units. In Proceedings of ACL 2016 (pp. 1715–1725).
13. Joshi, P., Santy, S., Budhiraja, A., Bali, K., & Choudhury, M. (2020). The state and fate of linguistic diversity and inclusion in the NLP world. In Proceedings of ACL 2020 (pp. 6282–6293).
14. Xolmanova, Z. T. (2024). Kompyuter lingvistikasi [Computational linguistics]. Tashkent.
15. Elov, B., Abdullayeva, O., & Primova, M. (2025). O‘zbek tilida tabiiy tilni qayta ishlashdagi morfologik, sintaktik va semantik tahlil metodlari [Morphological, syntactic and semantic analysis methods in Uzbek natural language processing].
16. Vaswani, A., et al. (2017). Attention is all you need. In Advances in Neural Information Processing Systems 30.


