From Algorithms to Pedagogy: A Web of Science-Based Bibliometric Study of Artificial Intelligence in English Reading Education

The swift advancement of artificial intelligence (AI) has reshaped English reading education by enabling automation, personalization, and intelligent feedback. This study examined the intellectual structure, thematic evolution, and emerging research fronts of AI in English reading education from 2021 to 2025. A bibliometric analysis of 279 peer-reviewed articles from the Web of Science (WoS) Core Collection was conducted using VOSviewer and CiteSpace. Results showed steady growth in research productivity and citation impact, with 2,091 citations and an H-index of 22. Bibliographic coupling identified three major intellectual streams: educational applications, technological algorithms, and cross-disciplinary extensions involving AI-enhanced readability, text complexity, and English for Medical Purposes. Keyword co-occurrence analysis revealed three thematic clusters focusing on AI-supported comprehension, deep-learning-driven machine reading comprehension (MRC) and natural language processing (NLP), and machine-learning-based readability research. Burst detection further indicated a shift from computational mechanisms such as “question answering” and “deep learning” toward pedagogical perspectives including “education,” “reading skills,” and “large language models.” Drawing on the Technology Acceptance Model (TAM), this study interpreted this transformation as a socio-technical process and provided transferable implications for responsible AI integration in English reading education.