Why does AI unlock new possibilities in STEM education? A Bibliometric Analysis of Trends and Future Agenda

Synthesis: STEM education faces challenges in personalization and interdisciplinary integration. AI technology has brought new possibilities, but the mechanisms by which AI reshapes the STEM education ecosystem require systematic investigation. This study employs bibliometric methods to analyze 242 publications from 2015-2025, constructing knowledge maps to reveal the evolutionary trajectory. The findings show that the field has transformed from intelligent tutoring systems to inquiry-based learning and computational thinking cultivation driven by LLMs. AI's key contribution lies in providing intelligent scaffolding that lowers the threshold for understanding knowledge. In this sense, AI is a core driving force promoting its shift from knowledge transmission to capability development.

Bibliometric analysis of 242 STEM-education papers (2015-2025) shows the field shifted from classic intelligent tutoring systems toward Large Language Models (LLMs)-driven inquiry-based learning and computational-thinking cultivation; AI's main value is intelligent scaffolding that lowers the knowledge threshold and moves STEM from knowledge transmission to capability development.

This work connects to core knowledge base themes: STEM Education Intelligent Tutoring Scaffolding Generative AI Adaptive Learning. It highlights how generative-AI tooling is reshaping both what learners do and how educators structure support, reinforcing the need for design that preserves authentic engagement rather than enabling shallow bypass.

What this means for practice

• Instructors. Position AI as intelligent Scaffolding that lowers the threshold for understanding content, and design tasks that use it to move learners from knowledge transmission toward capability development rather than recall.
• Researchers. Frame new work around the documented shift from intelligent tutoring systems to LLM-driven Inquiry-Based Learning and Computational Thinking cultivation, and treat the move from tool-oriented to system-oriented inquiry as the field's current direction.
• Policymakers. Fund integrated environments rather than isolated pilots: the analysis reports that early studies validated single technologies independently while current research emphasizes synergistic intelligent instructional environments.
• Instructors. Prepare for assessment to follow the same shift, since the personalization findings point toward dynamic, individually differentiated evaluation of competency pathways rather than standardized knowledge mastery.

Limitations

• This is a bibliometric study: keyword co-occurrence and theme evolution analysis with VOSviewer describe publication and keyword patterns, not learning outcomes, so it offers no causal evidence about AI's effectiveness.
• The analytical sample is 242 core documents screened from 2,146 Scopus records (2,125 after deduplication) published 2015–2025 in English; the authors note that limiting data sources to Scopus constrains the perspective.
• Findings rest on indexed titles, abstracts, and author-supplied keywords, so the maps reflect how researchers label their work rather than verified instructional designs.
• The screening step excluded non-research literature and studies judged of low relevance, a relevance judgment reported without inter-rater reliability statistics.

Citation

Chan, Chen, Hong, Song, Wang & Xu (2026). Why does AI unlock new possibilities in STEM education? A Bibliometric Analysis of Trends and Future Agenda. ISLS 2026 (arXiv preprint).