Mapping and Modeling the Role of Artificial Intelligence in Science Education: From Bibliometrics to Classroom Integration

This study presents a comprehensive and pedagogically grounded bibliometric and predictive investigation into the integration of artificial intelligence (AI) within science education. By applying advanced bibliometric techniques and predictive modeling, we examine global publication trends and emerging research themes between 2015 and 2024. Utilizing advanced regression methods—Linear Regression (LR) and Support Vector Regression (SVR)—we forecast scholarly activity, identify potential saturation points, and propose strategic alignments for future educational initiatives. Moving beyond bibliometric mapping, the study bridges AI research with pedagogical practice by proposing classroom-level implementation scenarios and structured teacher education frameworks. We demonstrate how generative AI tools (e.g., ChatGPT) can be embedded into science curricula to support inquiry-based learning, real-time feedback, and differentiated instruction. Furthermore, we outline how predictive modeling outcomes can inform teacher education programs focusing on ethical decision-making, AI tool selection, and culturally responsive AI-driven instruction. Our findings underscore the critical intersection of technological innovation and pedagogical preparedness. This multidimensional approach aims to empower researchers, educators, and policymakers to responsibly and effectively leverage AI technologies to advance science teaching and learning in diverse educational contexts worldwide.