Assessing the Intellectual Structure of the Evolving Knowledge Base on ChatGPT in the Field of Education and Health

Background/purpose –The unprecedented developments in AI-based technologies and large language models such as ChatGPT have exhibited a brand-new territory to be explored. Since its first release in November 2022, the potential utility of ChatGPT has garnered incremental attention in the scientific world, and has already accumulated a great number of studies from diverse fields. The current study was conducted with the purpose of exploring the scientific landscape of the evolving knowledge base related to the use of ChatGPT in the field of education and health through science mapping analysis of published research. Materials/methods – Data were retrieved from Web of Science and Scopus, and a comparative, period-based science mapping analysis was conducted using the SciMAT software. Results – The results showed that the studies published during the first period mostly focused on machine learning, reproductive medicine, education and first-year undergraduate themes. During the second period, though, the studies featured themes that are closely related to the design and performance of ChatGPT such as large language models (LLMs), natural language processing (NLP) and chatbot while abandoning a focus on artificial intelligence. These results imply that discussions and investigations over ChatGPT were being departed from those in the field of artificial intelligence, and the focus was becoming more central to the features of ChatGPT as a language model that can process huge amounts of information to generate human-like texts. Plagiarism and research ethics were also emerging themes during the last period. Conclusion – The results of the science mapping showed a growing interest into the opportunities and risks of ChatGPT, particularly for fields of education and medicine, and indicated that much research is warranted to discover the potential of GPT technology as an uncharted territory.