Bibliometric Analysis of the Use of Artificial Intelligence in Science Education

Abstract

Recent years have witnessed rapid advancements in artificial intelligence technologies, leading to significant transformations in science education. The integration of these technologies into teaching processes has garnered increasing attention for its potential to enhance the quality of learning experiences. In this context, this study aims to comprehensively analyze articles written at the intersection of science education and artificial intelligence using bibliometric methods.  Web of Science was selected as the database for this purpose. Citation, co-citation, bibliographic matching, co-authorship, and concept co-occurrence analyses were performed on the 147 articles obtained as a result of a three-stage screening process. These analyses were visualized using VOSviewer software. The study found that publications on science education and artificial intelligence showed a marked increase in 2024. The co-authorship analysis revealed that Zhai Xiaoming was the researcher with the highest connection power and collaborated with many countries. The United States, Germany, and China were identified as the countries with the most scientific contributions. At the institutional level, the University System of Georgia, the University of Georgia, and Michigan State University ranked in the top three in terms of research intensity. Key concept analysis revealed that the terms "Artificial Intelligence," "STEM Education," "Machine Learning," "ChatGPT," and "Science Education" were frequently encountered. The findings of this study systematically map AI-based articles written in recent years specifically on science education and provide a comprehensive perspective on publication dynamics, researcher profiles, and international collaborations in the field.