NLP METHODS FOR AUTOMATIC CLASSIFICATION OF VERBAL NON-FINITE FORMS IN THE UZBEK LANGUAGE
Keywords:
Uzbek language, natural language processing, NLP, non-finite verb forms, verbal noun, participle, converb, morphological analysis, POS tagging, Universal Dependencies, automatic classification.Abstract
This article investigates the automatic classification of non-finite verb forms in the Uzbek language, namely the verbal noun, participle, and converb, from the perspective of natural language processing (NLP). The agglutinative nature of Uzbek, the multifunctionality of morphological forms, and the context-dependent nature of grammatical meanings present specific challenges for automatic classification. The study analyzes rule-based, statistical, machine learning, and neural network approaches. It examines the possibility of representing Uzbek verb forms using the Fin, Part, Conv, and Vnoun values within the Universal Dependencies (UD) framework. Corpus resources developed for Uzbek, including Uzbek-UT, Uzbek-TueCL, UzUDT, and UzbekPOS, are analyzed, and a hybrid model for the automatic classification of non-finite verb forms is proposed based on these resources. The model integrates morphological features, affixal structures, and contextual information.
References
1. Elov, B. B., Khamroeva, Sh. M., Kurbanova, M. M., Khusainova, Z. Y., Ganiyeva, A. N. (2025). A Morphological Tagging Model of the Uzbek Language in the Universal Dependencies Format. 2025 10th International Conference on Computer Science and Engineering (UBMK), 380–384. DOI: 10.1109/UBMK67458.2025.11207015.
2. Universal Dependencies. UD for Uzbek. Universal Dependencies v2.
3. Universal Dependencies. Statistics of VERB in UD_Uzbek-UT.
4. Akhundjanova, A., Talamo, L. (2025). Universal Dependencies Treebank for Uzbek. Proceedings of the Third Workshop on Resources and Representations for Under-Resourced Languages and Domains (RESOURCEFUL-2025), 1–6.
5. Matlatipov, S. G., Aripov, M. (2026). UzUDT: Uzbek Universal Dependencies Treebank. Proceedings of the Fifteenth Language Resources and Evaluation Conference, 11642–11649. DOI: 10.63317/2uedjqezjxn5.
6. Sharipov, M., Kuriyozov, E., Vičič, J. (2026). UzbekPOS: A multi-domain dataset for Uzbek part-of-speech tagging. Data in Brief, 66, 112640. DOI: 10.1016/j.dib.2026.112640.
