Visualizing Research Trends on the Use of Generative AI in Assessment in the WOS database from 2019 to 2024 via Vosviewer and CiteSpace

The first typical generative Artificial Intelligence (Gen AI) model ChatGpt 1 was introduced in 2018, but has not drawn much attention until the release of ChatGpt 3 in 2022. Since then, researchers and educators have been experimenting with Gen AI tools to explore their possibilities in various fields. In education, the use of Gen AI in assessment is a research focus. To reveal the research patterns in this field, the study employed Vosviewer and CiteSpace to analyze 816 papers in the Web of Science (WOS) database published between 2019 and 2024. The annual publications surged during the period of 2023-2024 due to the release of higher versions of Gen AI tools such as ChatGpt 3. Researchers in the United States, United Kingdom and China engage most actively in the field. American and Hong Kong universities are particularly productive. However, the collaboration between institutions and authors still needs to be enhanced. Highly influential journals such as Nature and famous medical journals such as JMIR Medical Education and Cureus Journal of Medical sciences are most frequently cited. The analysis of co-occurrence keywords and keyword clusters identified two research areas: responding to the academic integrity issue with the use of Gen AI in assessment and exploring the valuable use of Gen AI in assessments in higher education especially medical education. Future research could explore the design of innovative or alternative assessments and the use of Gen AI tools in interactive and game-based assessment creation, marking and feedback giving.