Loading article...
Loading article...
Stream/Area:
-
Subject:
-
Paper ID:
IJSREM62978
Volume:
10 |
5 |
May 2026
Pages:
1-5
Publication Frequency:
Monthly (12 issues per year)
License:International Journal of Scientific Research in Engineering and Management follows the Creative Commons Attribution 4.0 (CC BY 4.0) License, allowing free use and distribution with proper attribution.
Department of Computer Science
Research Scholar, Vels Institute of Science, Technology & Advanced Studies (VISTAS), Chennai, India. Assistant Professor, Dwaraka Doss Goverdhan Doss Vaishnav College, Chenna
Department of Computer Science
Professor, Department of Information Technology & Advanced Studies (VISTAS), Chennai, India.,
Detecting and correcting spelling errors form an essential part of Natural Language Processing (NLP), as these processes significantly impact the accuracy and reliability of subsequent tasks, such as text classification, information retrieval, and machine translation. This study emphasizes the development of specialized spell-checking and correction systems tailored to specific domains. It examines various techniques ranging from traditional rule-based and hybrid approaches to modern deep learning architectures such as BiLSTM, CNN-LSTM, and Transformer-based models like BERT and BART, as well as large language models (LLMs). The evaluation relies on performance indicators including accuracy, precision, recall, F1-score, BLEU, and robustness. The results indicate that hybrid and deep learning methods deliver superior performance compared to conventional techniques, while LLMs and adversarial training approaches contribute to improved adaptability and stability. Moreover, the research highlights that low-resource languages, such as Tibetan, Wolof, Assamese, and Persian, gain notable advantages from these advanced models. Overall, the study underscores the effectiveness of contextual embeddings, bidirectional correction mechanisms, phonetic-based features, and multilingual zero-shot learning, suggesting that future efforts should focus on developing real-time, multilingual, and domain-adaptive solutions.
KEYWORDS
This survey analyzed 25 research works in domain-specific spelling correction using NLP. Deep learning and LLM-based models outperform traditional approaches, especially in handling domain-specific and low-resource languages. However, computational cost and efficiency remain challenges. Overall, NLP-based spell correction is moving toward real-time, multilingual, and domain-adaptive systems that can be applied in education, healthcare, legal, and technical fields. In the future, develop real-time spell checkers for social media and OCR data. Extend support for multilingual and low-resource languages. Build specialized correction systems for the medical and legal fields. Improve efficiency for large-scale applications. Integrate personalized feedback in education tools.
[1] A. Akın and S. Akın, “Zemberek: An open-source NLP framework for Turkish and other Turkic languages,” International Conference on Computational Linguistics, pp. 277–284, 2023. [2] H. Sun, L. Huang, and M. Li, “Query spelling correction using edit distance and context models,” IEEE Transactions on Knowledge and Data Engineering, vol. 36, no. 5, pp. 1154–1165, 2024. [3] M. Al-Khatib, R. Al-Khalil, and Y. Taha, “AraSpell: A deep learning approach for Arabic spelling correction,” Proceedings of the Association for Computational Linguistics (ACL), pp. 642–651, 2024. [4] J. Smith and R. Brown, “Improving English writing skills through NLP-driven error detection and correction systems,” International Journal of Educational Technology in Higher Education, vol. 20, no. 3, pp. 1–12, 2023. [5] A. Gupta and P. Kumar, “A data-driven approach for spellchecking and autocorrection,” IEEE Access, vol. 13, pp. 22541–22550, 2025. [6] T. Dorjee and S. Wangchuk, “TiSpell: A multilevel spelling correction system for Tibetan,” Journal of Computational Linguistics, vol. 49, no. 2, pp. 302–318, 2023. [7] Z. Liu, H. Wang, and Q. Chen, “Driving Chinese spelling correction through fine-grained evaluation,” Proceedings of the Conference on Empirical Methods in NLP (EMNLP), pp. 1212–1223, 2024. [8] N. Pham and V. Nguyen, “A BERT + Transformer approach for Vietnamese spelling correction,” IEEE Transactions on Computational Linguistics, vol. 11, no. 1, pp. 23–35, 2024. [9] P. Zhang and K. Zhou, “SAGE: Generative spelling correction with automatic error emulation,” Proceedings of NAACL-HLT, pp. 511–522, 2025. [10] M. Diop and C. Fall, “Transformer-based spell checking for Wolof: A low-resource language study,” International Conference on African Language Technology (AfriNLP), pp. 94–103, 2023.
Publication Date
5/9/2026
Peer Review Type
Double-Blind Peer Review
Review Rounds: 2
Documents
Article Views
0
Total views since publication
PDF Downloads
3
Total downloads since publication
Geographic Reach
5
Digital Preservation: This article is permanently archived on Zenodo and complies with all data preservation standards across major indexing platforms.
Plagiarism Screening
This manuscript has been checked for plagiarism using TurnitinPlagiarismChecker software.
Conflict of Interest
The authors declare no conflict of interest related to this research work.
Funding Information
No funding information available for this research.
Related research in the same subject area
Loading similar articles...