Why Does AI Unlock New Possibilities in STEM Education: A Bibliometric Analysis of Trends and Future Agenda

Jul 16, 2026

Upload by Aria

STEM education faces persistent difficulties in personalization and in integrating interdisciplinary learning. Although AI introduces new opportunities, the processes through which AI reshapes the STEM education ecosystem need systematic study. This research applies bibliometric methods to 242 publications from 2015–2025 and builds knowledge maps to trace the field’s evolution. Results indicate a shift from intelligent tutoring systems toward inquiry-based learning and computational thinking supported by LLMs. AI’s central value is intelligent scaffolding that lowers barriers to understanding, enabling movement from knowledge transmission to capability development.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages6

Reading Time15 min

Abstract

Introduction

Methodology and materials

Research methodology

Data source and filtering

Bibliometric results and analysis

Trends in publications and disciplinary distribution

Research hotspots and evolutionary trends

Why does AI unlock new possibilities in STEM education? ABibliometric Analysis of Trends and Future Agenda Jesse Yusuf Chan(Zexi Chen), East China Normal University, [51274108057@stu.ecnu.edu.cn](51274108057@stu.ecnu.edu.cn)[ ](51274108057@stu.ecnu.edu.cn)Mengyao Chen, East China Normal University, [51274108033@stu.ecnu.edu.cn](51274108033@stu.ecnu.edu.cn)[ ](51274108033@stu.ecnu.edu.cn)Yang Hong, East China Normal University, [HongYang387@outlook.com](HongYang387@outlook.com)[ ](HongYang387@outlook.com)Ziyun Song, Nantong University, [drivingintosea@gmail.com](drivingintosea@gmail.com)[ ](drivingintosea@gmail.com)Haoming Wang*, East China Normal University, [wfrank0222@gmail.com](wfrank0222@gmail.com)[ ](wfrank0222@gmail.com)Xianlong Xu*, East China Normal University, [xlxu@eec.ecnu.edu.cn](xlxu@eec.ecnu.edu.cn) Abstract: STEM education faces challenges in personalization and interdisciplinary integration. AI technology has brought new possibilities, but the mechanisms by which AIreshapes 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. Introduction STEM education is undergoing a profound transformation driven by AI technology. In particular, breakthroughsin large language models (LLMs) demonstrate significant potential in optimizing instructional design, constructing personalized learning environments, and cultivating critical thinking and problem-solving skills (Lin et al., 2025) . However, challenges remain in integrating technology with curriculum, restructuring pedagogical models, and cultivating emerging AI literacy among both teachers and students (Yang et al., 2025) . Existing reviews have explored chatbots in scientific computing education (Groothuijsen et al., 2024), AI teaching tools for secondary school students (Crompton et al., 2024), and AI for interdisciplinary learning (Cai et al., 2025) . These studies often focus on qualitative analysis or a single technological application, lacking a macroscopic knowledge map of AI across STEM education. Therefore, this study adopts a bibliometric approach to analyze core literature from the Scopus database between 2015 and 2025, aiming to answer the following research questions: RQ1: What are the current core research hotspots and directions for AI in the field of STEM education? RQ2: How has research output on AI in STEM education evolved over the past decade? Methodology and materials Research methodology As a scientific research method, bibliometrics utilizes metrical concepts and tools to analyze literature data (Jinget al., 2024) . This method can objectively track the developmental trajectory of a discipline, predict future research directions, and provide important references for academic planning (Wang et al., 2025a) . This study employs the VOSviewer software for data analysis, conducting keyword co-occurrence analysis and research theme evolution analysis on literature in the field of AI in STEM education. Data source and filtering The literature search was conducted in the Scopus database. The search query was: TITLE-ABS-KEY(steam OR stem) AND TITLE-ABS-KEY(AI OR AI-Agent OR GPT OR LLM) AND TITLE-ABS-KEY(educat OR learn OR teach OR instruct) . The time span was set from 2015 to 2025, limited to English-language publications, yielding 2, 146 documents (2, 125 after deduplication) . The literature was screened by excluding non-research literature and studies with low relevance to the r

FAQs

What core challenges in STEM education does the study highlight as targets for AI support?

The study highlights personalization difficulties and weak interdisciplinary integration in STEM education, and argues that AI may address these gaps.

How does the study conduct its bibliometric analysis?

The research uses bibliometrics with VOSviewer, performing keyword co-occurrence analysis and theme evolution analysis on Scopus-indexed literature.

What does the study find about how AI research in STEM education has evolved over time?

Publication output rises steadily from 2015 onward, accelerates after 2019, and shows explosive growth after 2022 with widespread LLM implementation, indicating sustained future growth.

Jul 16, 2026

Upload by Aria