Multi-granularity Curriculum Learning for Chinese Spelling Correction in Legal Texts
DOI:
https://doi.org/10.15837/ijccc.2026.3.7213Keywords:
Chinese Spelling Correction, Curriculum Learning, Legal TextsAbstract
Chinese spelling correction (CSC) in legal texts faces unique challenges due to specialized terminology and complex error patterns. To address this gap, we propose Multi-granularity Curriculum Learning (mgCL), a novel two-level adaptive training framework that integrates batch-level and instance-level curricula. At the batch level, mgCL dynamically prioritizes training samples based on cross-entropy-derived difficulty, ensuring the model is exposed to increasingly complex examples in line with its evolving competence. At the instance level, mgCL leverages Monte Carlo Dropout to quantify prediction uncertainty and focus correction on challenging characters, particularly domain-specific legal terms. To facilitate evaluation in this domain, we also introduce CNLAW, a novel benchmark dataset for legal-domain CSC, featuring diverse error patterns and extensive legal terminology. Experiments present the effectiveness of our framework: on CNLAW, mgCL achieves 98.02% F1, outperforming the strong Rephrasing Language Model (ReLM) baseline (96.75%), and dramatically reduces the False Positive Rate (FPR) from 1.60% to 0.16%. Moreover, robust performance on the widely used SIGHAN15 benchmark confirms mgCL’s cross-domain generalization.
References
Yu, J.; Li, Z. (2014). Chinese spelling error detection and correction based on language model, pronunciation, and shape, Proceedings of The Third CIPS-SIGHAN Joint Conference on Chinese Language Processing, 220-223, 2014. https://doi.org/10.3115/v1/W14-6835
Wang, D.; Song, Y.; Li, J.; Han, J.; Zhang, H. (2018). A hybrid approach to automatic corpus generation for Chinese spelling check, Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, 2517-2527, 2018. https://doi.org/10.18653/v1/D18-1273
Yu, L.-C.; Lee, L.-H.; Tseng, Y.-H.; Chen, H.-H. (2014). Overview of SIGHAN 2014 bake-off for Chinese spelling check, Proceedings of The Third CIPS-SIGHAN Joint Conference on Chinese Language Processing, 126-132, 2014. https://doi.org/10.3115/v1/W14-6820
Yu, L.-C.; Lee, L.-H.; Tseng, Y.-H.; Chen, H.-H. (2014). Overview of SIGHAN 2014 bake-off for Chinese spelling check, Proceedings of The Third CIPS-SIGHAN Joint Conference on Chinese Language Processing, 126-132, 2014. https://doi.org/10.3115/v1/W14-6820
Wang, X. et al. (2024). An empirical investigation of domain adaptation ability for chinese spelling check models, ICASSP 2024-2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), IEEE, 2024. https://doi.org/10.1109/ICASSP48485.2024.10448189
Wang, X. et al. (2024). An Unsupervised Domain-Adaptive Framework for Chinese Spelling Checking, ACM Transactions on Asian and Low-Resource Language Information Processing, 23(11): 1-16, 2024. https://doi.org/10.1145/3689821
Hong, Y.; Yu, X.; He, N.; Liu, N.; Liu, J. (2019). FASPell: A fast, adaptable, simple, powerful Chinese spell checker based on DAE-decoder paradigm, Proceedings of the 5th Workshop on Noisy User-generated Text (W-NUT 2019), 160-169, 2019. https://doi.org/10.18653/v1/D19-5522
Wang, B.; Che, W.; Wu, D.; Wang, S.; Hu, G.; Liu, T. (2021). Dynamic connected networks for chinese spelling check, Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021, 2437-2446, 2021. https://doi.org/10.18653/v1/2021.findings-acl.216
Bai, J.; Bai, S.; Chu, Y. et al. (2023). Qwen technical report, arXiv preprint arXiv:2309.16609, 2023.
Achiam, J.; Adler, S.; Agarwal, S. et al. (2023). Gpt-4 technical report, arXiv preprint arXiv:2303.08774, 2023.
Liu, S.; Yang, T.; Yue, T.; Zhang, F.; Wang, D. (2021). PLOME: Pre-training with misspelled knowledge for Chinese spelling correction, Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers), 2991-3000, 2021. https://doi.org/10.18653/v1/2021.acl-long.233
Wu, H.; Zhang, H.; Xuan, R.; Song, D. (2024). Bi-DCSpell: A bidirectional detector-corrector interactive framework for Chinese spelling check, Findings of the Association for Computational Linguistics: EMNLP 2024, 3974-3984, 2024. https://doi.org/10.18653/v1/2024.findings-emnlp.229
Liu, L.; Wu, H.; Zhao, H. (2024). Chinese Spelling Correction as Rephrasing Language Model, Proceedings of the AAAI Conference on Artificial Intelligence, 38(17): 18662-18670, 2024. https://doi.org/10.1609/aaai.v38i17.29829
Bengio, Y.; Louradour, J.; Collobert, R. et al. (2009). Curriculum learning, Proceedings of the 26th annual international conference on machine learning, 41-48, 2009. https://doi.org/10.1145/1553374.1553380
Gan, Z.; Xu, H.; Zan, H. (2021). Self-supervised curriculum learning for spelling error correction, Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, 3487- 3494, 2021. https://doi.org/10.18653/v1/2021.emnlp-main.281
Zhou, Y.; Yang, B.; Wong, D. F.; Wan, Y.; Chao, L. S. (2020). Uncertainty-aware curriculum learning for neural machine translation, Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, 6934-6944, 2020. https://doi.org/10.18653/v1/2020.acl-main.620
Li, J.; Wang, Q.; Mao, Z.; Guo, J.; Yang, Y.; Zhang, Y. (2022). Improving Chinese spelling check by character pronunciation prediction: The effects of adaptivity and granularity, Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, 4275-4286, 2022. https://doi.org/10.18653/v1/2022.emnlp-main.287
Huang, H.; Ye, J.; Zhou, Q.; Li, Y.; Li, Y.; Zhou, F.; Zheng, H.-T. (2023). A frustratingly easy plug-and-play detection-andreasoning module for Chinese spelling check, Findings of the Association for Computational Linguistics: EMNLP 2023, 11514-11525, 2023. https://doi.org/10.18653/v1/2023.findings-emnlp.771
Liang, Z.; Quan, X.; Wang, Q. (2023). Disentangled phonetic representation for Chinese spelling correction, Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics, 13509-13521, 2023. https://doi.org/10.18653/v1/2023.acl-long.755
Huang, L.; Li, J.; Jiang, W.; Zhang, Z.; Chen, M.; Wang, S.; Xiao, J. (2021). Phmospell: Phonological and morphological knowledge guided chinese spelling check, Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing, 5958-5967, 2021. https://doi.org/10.18653/v1/2021.acl-long.464
Cheng, X.; Xu, W.; Chen, K.; Jiang, S.; Wang, F.; Wang, T.; Chu, W.; Qi, Y. (2020). Spellgcn: Incorporating phonological and visual similarities into language models for chinese spelling check, Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, 871-881, 2020. https://doi.org/10.18653/v1/2020.acl-main.81
Liu, C.; Zhang, K.; Jiang, J.; Liu, Z.; Tao, H.; Gao, M.; Chen, E. (2024). ARM: An alignment-andreplacement module for Chinese spelling check based on LLMs, Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, 10156-10168, 2024. https://doi.org/10.18653/v1/2024.emnlp-main.567
Xu, H.-D.; Li, Z.; Zhou, Q.; Li, C.; Wang, Z.; Cao, Y.; Huang, H.; Mao, X.-L. (2021). Read, listen, and see: Leveraging multimodal information helps Chinese spell checking, Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021, 716-728, 2021. https://doi.org/10.18653/v1/2021.findings-acl.64
Zhang, S.; Huang, H.; Liu, J.; Li, H. (2020). Spelling error correction with soft-masked BERT, Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, 882-890, 2020. https://doi.org/10.18653/v1/2020.acl-main.82
Zhu, C.; Ying, Z.; Zhang, B.; Mao, F. (2022). Mdcspell: A multi-task detector-corrector framework for chinese spelling correction, Findings of the Association for Computational Linguistics: ACL 2022, 1244-1253, 2022. https://doi.org/10.18653/v1/2022.findings-acl.98
Huang, H.; Ye, J.; Zhou, Q.; Li, Y. (2023). A Frustratingly Easy Plug-and-Play Detection-and- Reasoning Module for Chinese Spelling Check, Findings of the Association for Computational Linguistics: EMNLP 2023, 11514-11525, 2023. https://doi.org/10.18653/v1/2023.findings-emnlp.771
Wu, H.; Zhang, H.; Xuan, R.; Song, D. (2024). Bi-DCSpell: A Bi-directional Detector-Corrector Interactive Framework for Chinese Spelling Check, findings of the Association for Computational Linguistics: EMNLP 2024, 3974-3984, 2024. https://doi.org/10.18653/v1/2024.findings-emnlp.229
Sarafianos, N.; Giannakopoulos, T.; Nikou, C.; Kakadiaris, I. A. (2018). Curriculum learning of visual attribute clusters for multi-task classification, Pattern Recognition, 94-108, 2018. https://doi.org/10.1016/j.patcog.2018.02.028
Wang, Y.; Gan, W.; Yang, J.; Wu, W.; Yan, J. (2019). Dynamic Curriculum Learning for Imbalanced Data Classification, Proceedings of IEEE/CVF International Conference on Computer Vision, 5016-5025, 2019. https://doi.org/10.1109/ICCV.2019.00512
Guo, S.; Huang, W.; Zhang, H.; Zhuang, C.; Dong, D.; Scott, M. R.; Huang, D. (2018). CurriculumNet: Weakly Supervised Learning from Large-Scale Web Images, Proceedings of Computer Vision - ECCV 2018: 15th European Conference, 139-154, 2018. https://doi.org/10.1007/978-3-030-01249-6_9
Cirik, V.; Hovy, E.; Morency, L. P. (2016). Visualizing and understanding curriculum learning for long short-term memory networks, arXiv preprint arXiv:1611.06204, 2016.
Liu, C.; He, S.; Liu, K.; Zhao, J. (2018). Curriculum learning for natural answer generation, Proceedings of the 27th International Joint Conference on Artificial Intelligence, 4223-4229, 2018. https://doi.org/10.24963/ijcai.2018/587
Kocmi, T.; Bojar. (2017). Curriculum Learning and Minibatch Bucketing in Neural Machine Translation, Proceedings of the International Conference Recent Advances in Natural Language Processing, 379-386, 2017. https://doi.org/10.26615/978-954-452-049-6_050
Thompson, B.; Khayrallah, H.; Anastasopoulos, A.; McCarthy, A. D.; Duh, K.; Marvin, R.; McNamee, P.; Gwinnup, J.; Anderson, T.; Koehn, P. (2018). Freezing Subnetworks to Analyze Domain Adaptation in Neural Machine Translation, Proceedings of the Third Conference on Machine Translation: Research Papers, 124-132, 2018. https://doi.org/10.18653/v1/W18-6313
Zhang, X.; Shapiro, P.; Kumar, G.; McNamee, P.; Carpuat, M.; Duh, K. (2019). Curriculum Learning for Domain Adaptation in Neural Machine Translation, Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, 1903-1915, 2019. https://doi.org/10.18653/v1/N19-1189
Wang, W.; Caswell, I.; Chelba, C. (2019). Dynamically Composing Domain-Data Selection with Clean-Data Selection by "Co-Curricular Learning" for Neural Machine Translation, Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, 1282-1292, 2019. https://doi.org/10.18653/v1/P19-1123
Kumar, G.; Foster, G.; Cherry, C.; Krikun, M. (2019). Reinforcement Learning Based Curriculum Optimization for Neural Machine Translation, Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics, 2054-2061, 2019. https://doi.org/10.18653/v1/N19-1208
Zhang, X.; Kumar, G.; Khayrallah, H.; Murray, K.; Gwinnup, J.; Martindale, M. J.; McNamee, P.; Duh, K.; Carpuat, M. (2018). An Empirical Exploration of Curriculum Learning for Neural Machine Translation, CoRR, abs/1811.00739, 2018.
Platanios, E. A.; Stretcu, O.; Neubig, G.; Poczos, B.; Mitchell, T. (2019). Competence-Based Curriculum Learning for Neural Machine Translation, Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologiesm, 1162-1172, 2019. https://doi.org/10.18653/v1/N19-1119
Additional Files
Published
Issue
Section
License
Copyright (c) 2026 Shijin Zhou, Yabo Liu
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
ONLINE OPEN ACCES: Acces to full text of each article and each issue are allowed for free in respect of Attribution-NonCommercial 4.0 International (CC BY-NC 4.0.
You are free to:
-Share: copy and redistribute the material in any medium or format;
-Adapt: remix, transform, and build upon the material.
The licensor cannot revoke these freedoms as long as you follow the license terms.
DISCLAIMER: The author(s) of each article appearing in International Journal of Computers Communications & Control is/are solely responsible for the content thereof; the publication of an article shall not constitute or be deemed to constitute any representation by the Editors or Agora University Press that the data presented therein are original, correct or sufficient to support the conclusions reached or that the experiment design or methodology is adequate.
