Optimizing Prompt Engineering in Translation Practice: A Comparative Study of ChatGPT-4.0 and ChatGPT-4.o Mini

Wei WANG, Weihong ZHOU

Abstract


Recent advancements in natural language processing (NLP) have introduced large language models (LLMs) like ChatGPT-4.0 and its smaller variant, ChatGPT-4o mini, which are increasingly utilized in machine-aided translation (MT). This study investigates the critical role of prompt engineering in enhancing translation quality using these models. By conducting a systematic comparison between ChatGPT-4.0 and ChatGPT-4o mini, we examine how different prompt designs influence translation accuracy and fluency across various text types. Through rigorous literature review, empirical analysis, and comprehensive sample evaluation, this research provides an in-depth assessment of how prompt engineering affects translation outputs. The findings offer practical recommendations for improving translation practices and set the stage for future research in this evolving domain.


Keywords


Prompt engineering; Machine-aided translation; ChatGPT-4.0; ChatGPT-4o Mini; Natural language processing; Translation quality; Language models

Full Text:

PDF

References


Bahdanau, D., Cho, K., & Bengio, Y. (2014). Neural machine translation by jointly learning to align and translate. arXiv, Article 1409.0473.

Brown, T., Mann, B., Ryder, N., Subbiah, M., Kaplan, J., Dhariwal, P., et al. (2020). Language Models Are Few-Shot Learners. arxiv preprint arxiv:2005.14165.

Gao, Z., Liu, Y., & Zhao, Y. (2022). Enhancing Translation Models via Prompt Engineering: A Survey. Journal of Machine Translation, 15(3), 210-224.

Johnson, A., Keller, M., & Wu, Y. (2022). A Study on Prompt Engineering in Machine Translation Models. AI Journal, 9(2), 45-63.

Liu, S., Zhang, L., & Wang, M. (2023). Optimizing Large Language Models for Multilingual Translation. Computational Linguistics Review, 14(1), 65-78.

OpenAI. (2023). GPT-4 Technical Report. OpenAI.

OpenAI. (2024). ChatGPT-4.0 Mini Technical Overview. OpenAI.

Vaswani, A., Shazeer, N., Parmar, N., et al. (2017). Attention is All You Need. Advances in Neural Information Processing Systems, 30, 5998-6008.

Xia, P., Liu, Z., & Tang, J. (2023). Prompting Neural Models for Enhanced Machine Translation. Proceedings of the ACL 2023 Conference, 483-493.

Yamada, M. (2023). Optimizing Machine Translation through prompt engineering: An Investigation into ChatGPT’s Customizability. Proceedings of Machine Translation Summit XIX (September 4-8, 2023, Macau SAR, China.), Vol. 2: Users Track, pp.195-204.




DOI: http://dx.doi.org/10.3968/13573

Refbacks

  • There are currently no refbacks.


Copyright (c) 2024 Canadian Social Science

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 International License.

Reminder

  • How to do online submission to another Journal?
  • If you have already registered in Journal A, then how can you submit another article to Journal B? It takes two steps to make it happen:

Submission Guidelines for Canadian Social Science

We are currently accepting submissions via email only. The registration and online submission functions have been disabled.

Please send your manuscripts to css@cscanada.net，or css@cscanada.org for consideration. We look forward to receiving your work.

 Articles published in Canadian Social Science are licensed under Creative Commons Attribution 4.0 (CC-BY).

 

Canadian Social Science Editorial Office

Address: 1020 Bouvier Street, Suite 400, Quebec City, Quebec, G2K 0K9, Canada.
Telephone: 1-514-558 6138 
Website: Http://www.cscanada.net; Http://www.cscanada.org 
E-mail:caooc@hotmail.com; office@cscanada.net

Copyright © Canadian Academy of Oriental and Occidental Culture