Evaluating ChatGPT’s Multilingual Performance in Clinical Nutrition Advice Using Synthetic Medical Text: Insights from Central Asia
Our findings show that, despite using identical prompts across 3 distinct languages, the ChatGPT-4's capability to produce sensible outputs is limited by the lack of training data in non-English languages. Thus, a customized large language model should be developed to perform better in underrepresented languages and to take into account specific local diets and practices.
Authors
- Hüseyin Atakan Varol (ORCID: https://orcid.org/0000-0002-4042-425X)
- Aknur Karabay (ORCID: https://orcid.org/0000-0001-6160-6661)
- Aizhan Meyerbekova (ORCID: https://orcid.org/0009-0001-8260-913X)
- Mei‐Yen Chan (ORCID: https://orcid.org/0000-0003-2324-9338)
- Gulnoza Adilmetova (ORCID: https://orcid.org/0009-0006-0836-4767)
- Ruslan Nassyrov
Institutions
- Nazarbayev University (KZ)
Publication Details
- Journal
- Journal of Nutrition
- Published
- 2024-12-26
- DOI
- https://doi.org/10.1016/j.tjnut.2024.12.018
- Citations
- 11
- Primary Topic
- Artificial Intelligence in Healthcare and Education
- Type
- article
- Field-Weighted Citation Impact
- 1.32
Funders
- Ministry of Education and Science of the Republic of Kazakhstan
- Nazarbayev University