ArticleTranslation Theories and Skills http://dx.doi.org/10.26855/tfll.2025.12.010
A Preliminary Study on the Translation Quality of Legal Texts by Large Language Models Based on Parallel Corpus
Xiaoyang Zhang, Hongjing Dong*
Slavic Language Institute, Harbin Normal University, Harbin 150025, Heilongjiang, China.
*Corresponding author: Hongjing Dong
2024 Heilongjiang Provincial Philosophy and Social Science Research Planning Project, Research on the Translation Quality of Russian Translation of the Civil Code of the People’s Republic of China (Project No. 24YYC002); 2025 Heilongjiang Provincial College Students’ Innovative Training Program Project, Research on Translation Quality Evaluation and Enhancement Strategies for Translating Legal Texts in Big Language Models Based on Parallel Corpora (Project no: S202510231098).
Published: December 31, 2025
Abstract
Compared with traditional human translation, large language model shows obvious advantages in terms of efficiency and cost, but its stability, terminological consistency and standardized expression in the legal domain still have shortcomings. In this paper, we construct a Chinese-Russian parallel corpus of the General Provisions of the First Subpart of the Contract Part of the Civil Code of the People’s Republic of China (hereinafter referred to as the Civil Code), quantitatively compare the differences between the direct translation results of the large lan-guage model of DeepSeek, ChatGPT, and Yandex and the professional human translations by using the corpus, and accurately reveal the limitations of the large language model in the translation of specific legal semantic fields (e.g., terminology, syntax, and logic), and explore and propose optimization strategies, ultimately aiming to provide data support and practical references for optimizing the application of large language models in the field of legal translation.
Keyword
Comparison of human-computer translations; parallel corpus; large language model; Civil Code of the People’s Republic of China
References
Gao, H. R. (1993). The usage of ДОГОВОР and КОНТРАКТ. Foreign Languages Research, (4), 53-54.
Li, F. X., Zhang, Y., & Ding, L. J. (2025). A comparison of translation quality between large language models and neural machine translation systems in specialized texts: A case study of Chinese-English legal translation. Shanghai Journal of Translators, (6), 62-67.
Shi, Y. Q., Wang, M. J., & Xu, J. Y. (2025). Causes and countermeasures of terminology mistranslation with the assistance of large language models. China Terminology, 27(5), 36-43.
Song, L. J. (2024). A study on the digital humanities transformation of legal translation: Centering on specialized data-bases and ChatGPT. Foreign Language Research, (2), 51-57.
Wu, C. H., & Zou, J. T. (2025). Challenges and optimization of machine translation for legal terminology: Terminology generation mechanisms and human-computer collaboration strategies. Shanghai Journal of Translators, (6), 56-61.
Zhang, F. L. (2020). On machine translation technology in legal translation. Technology Enhanced Foreign Languages, (1), 53-58+8.
Zhang, H. C. (2024). A study on the reform of Russian arbitration institutions [Unpublished doctoral dissertation]. Heilongjiang University.
Zhao, J. F., & Li, X. (2024). Construction and application of translation agents driven by large language models. Technology Enhanced Foreign Languages, (5), 22-28+75+108.
Copyright
© 2025 by the author(s).
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution-NonCommercial-NoDerivatives (CC BY-NC-ND) license, which permits non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited and is not modified or adapted.
https://creativecommons.org/licenses/by-nc-nd/4.0/
How to cite this paper
A Preliminary Study on the Translation Quality of Legal Texts by Large Language Models Based on Parallel Corpus
How to cite this paper: Xiaoyang Zhang, Hongjing Dong. (2025). A Preliminary Study on the Translation Quality of Legal Texts by Large Language Models Based on Parallel Corpus. Translation and Foreign Language Learning, 1(5), 806-810.
DOI: http://dx.doi.org/10.26855/tfll.2025.12.010