1 Department of Computational Linguistics and Digital Technologies, Tashkent State University of Uzbek Language and Literature
ReceivedOct 28, 2026
RevisedNov 12, 2026
PublishedNov 17, 2026
Abstract
Automatic spelling correction in Uzbek must address productive agglutination, Latin–Cyrillic variation, apostrophe-like signs, proper names, borrowings, and context-dependent real-word errors. This conceptual and technical study proposes a morphology- and context-aware hybrid framework that separates orthographic normalization, lexical and morphological analysis, error detection, candidate generation, contextual ranking, and confidence-based acceptance or abstention. A selective narrative synthesis of 17 works compares approaches by error coverage, morphological suitability, context dependence, resource requirements, interpretability, and overcorrection risk. The proposed candidate score combines normalized contextual likelihood, weighted edit cost, morphological plausibility, and error-channel evidence. Parameter selection uses validation-only tuning, morphological ambiguity aggregation, calibrated confidence, and an explicit no-change candidate. Uzbek-specific policies include reversible apostrophe normalization, script-aware handling, suffix-sequence constraints, contextual screening of valid words, and conservative protection of unknown names and technical terms. A prospective evaluation protocol distinguishes natural from synthetic errors, prevents document-level data leakage, and measures candidate Recall@k, detection precision, recall and F1, correction accuracy, overcorrection, calibration, and runtime. Proposed ablations isolate morphology, context, error-channel modeling, and abstention. No corpus construction, implementation results, performance gains, or statistically significant superiority is claimed. The contribution is a testable technical specification for subsequent Uzbek-language implementation and empirical validation.
1. Bhaire, V. V., Jadhav, A. A., Pashte, P. A., & Magdum, P. G. (2015). Spell checker. International Journal of Scientific and Research Publications, 5(4), 1–3. https://www.ijsrp.org/research-paper-0415/ijsrp-p4076.pdf
2. Downs, B. (2020). Improving spellchecking for children: Correction and design [Master’s thesis, Boise State University]. Boise State University ScholarWorks. https://scholarworks.boisestate.edu/td/1721/
3. Elov, B. B., Adalı, E., Khamroeva, S. M., Abdullayeva, O. X., Xusainova, Z. Y., & Xudayberganov, N. U. (2023). The problem of POS tagging and stemming for agglutinative languages (Turkish, Uyghur, Uzbek languages). In 2023 8th International Conference on Computer Science and Engineering (UBMK) (pp. 57–62). IEEE. https://doi.org/10.1109/UBMK59864.2023.10286792
4. Elov, B. B., Hamroyeva, S. M., & Axmedova, X. I. (2023). Methods for creating a morphological analyzer. In H. Zaynidinov, M. Singh, U. S. Tiwary, & D. Singh (Eds.), Intelligent human computer interaction (Lecture Notes in Computer Science, Vol. 13741, pp. 27–38). Springer. https://doi.org/10.1007/978-3-031-27199-1_4
5. Elov, B. B., Khamroeva, S. M., & Xusainova, Z. Y. (2023). The pipeline processing of NLP. E3S Web of Conferences, 413, Article 03011. https://doi.org/10.1051/e3sconf/202341303011
6. Elov, B. B., Sirojiddinov, S. S., Khamroeva, S. M., Adalı, E., & Xusainova, Z. Y. (2023). Pos taging of Uzbek text using hidden Markov model. In 2023 8th International Conference on Computer Science and Engineering (UBMK) (pp. 63–68). IEEE. https://doi.org/10.1109/UBMK59864.2023.10286619
7. Elov, B. B., Xusainova, Z. Y., Umirova, S. M., Normamatov, S., Abdimurod kızı, A. S., & Mahmadiev, S. (2024). Algorithms for parsing roots and stems of words in Uzbek language. In 2024 9th International Conference on Computer Science and Engineering (UBMK). IEEE. https://doi.org/10.1109/UBMK63289.2024.10773598
8. Gorin, R. E. (1974, December 14). SPELL: Spelling check and correction program [Program documentation]. Stanford Artificial Intelligence Laboratory. https://www.saildart.org/SPELL.REG%5BUP,DOC%5D3
9. Jurafsky, D., & Martin, J. H. (2024). Speech and language processing (3rd ed. draft, August 20, Chapter 3, “N-gram language models”). https://web.stanford.edu/~jurafsky/slp3/old_aug24/3.pdf
10. Kumar, R., Bala, M., & Sourabh, K. (2018). A study of spell checking techniques for Indian languages. JK Research Journal in Mathematics and Computer Sciences, 1(1), 105–113. https://jkhighereducation.nic.in/jkrjmcs/issue1/15.pdf
11. Lavoshnikova, E. K. (2015). Proverka orfografii v tekstovom redaktore Word 2013: Imena sobstvennye [Spell check in Word 2013: Proper names]. Aktualnye voprosy obshchestvennykh nauk, (8).
12. Lavoshnikova, E. K. (2017). “Problemnye” slova kak prichina propuska oshibok pri kompyuternoy proverke orfografii [“Problematic” words as a cause of missed errors in computer spell checking]. Tekst. Kniga. Knigoizdanie, (15), 113–129. https://doi.org/10.17223/23062061/15/8
13. Murodov, S. A., & Xazratqulov, S. Z. (2022). Spellchecker Windows ilovasi orqali o‘zbek tilidagi imloviy savodxonlikni tekshirish usullari [Methods for checking Uzbek spelling literacy through the Spellchecker Windows application]. Academic Research in Educational Sciences, 3(6), 364–369. https://cyberleninka.ru/article/n/spellchecker-windows-ilovasi-orqali-o-zbek-tilidagi-imloviy-savodxonlikni-tekshirish-usullari
14. Pirinen, T. A., & Lindén, K. (2010). Finite-state spell-checking with weighted language and error models: Building and evaluating spell-checkers with Wikipedia as corpus. In Proceedings of LREC 2010: Workshop on creation and use of basic lexical resources for less-resourced languages. https://helda.helsinki.fi/server/api/core/bitstreams/3d9eddbe-e4d7-4a39-9159-30d1566cafd3/content
15. Richter, M., Straňák, P., & Rosen, A. (2012). Korektor: A system for contextual spell-checking and diacritics completion. In Proceedings of COLING 2012: Posters (pp. 1019–1028). The COLING 2012 Organizing Committee. https://aclanthology.org/C12-2099/
16. Sharma, S., Valls-Vargas, J., King, T. H., Guerin, F., & Arora, C. (2023). Contextual multilingual spellchecker for user queries. In Proceedings of the 46th International ACM SIGIR Conference on Research and Development in Information Retrieval (pp. 3395–3399). Association for Computing Machinery. https://doi.org/10.1145/3539618.3591861
How to Cite
Sobirova Nazira (2026). AUTOMATIC SPELLING ERROR DETECTION AND CORRECTION IN UZBEK TEXTS: A MORPHOLOGY- AND CONTEXT-AWARE HYBRID FRAMEWORK. Science and Innovation, 5(10), 25-34. https://doi.org/10.67620/sai.510c0004