ChatGPT is changing the way we test language more and more by using advanced deep learning models that mimic how people talk to each other. These models, especially large language models (LLMs), can understand, create, and evaluate natural language. This makes them very useful for educational evaluation. This report looks at how ChatGPT is changing the way we evaluate things by looking at academic articles from 2023 to 2024. We used the Web of Science database to carefully choose 350 peer-reviewed articles that looked at worldwide research trends. Bibliometric technologies like VOSviewer and Biblioshiny were used to find important changes in the field. These technologies made it possible to see how publication patterns, co-authorship networks, and the development of thematic trends relating to the function of ChatGPT in education have changed over time. The results showed a number of important things. First, the United States, China, and the United Kingdom were the top contributors to this sector. This shows that institutions and governments in these countries are very interested in using AI language models in education. Second, Elyoseph Z. and Levkovich I. were two of the most numerous writers who contributed to the discussion around AI-driven assessment. Some of the most well-known journals that published this research are JMIR Medical Education, Journal of Medical Internet Research, JMIR Formative Research, and Education Sciences. The keyword co-occurrence analysis showed that words like "Chatgpt," "assessment," and "artificial intelligence" are used a lot, which strengthens the connection between computational linguistics and educational evaluation.
Sabirova et al. (Mon,) studied this question.