A Bibliometric and LDA Topic Modeling Analysis of Artificial Intelligence in English Language Learning

Abstract

Artificial Intelligence (AI) has become a transformative tool in education, particularly in the domain of English Language Learning (ELL), offering adaptive, personalized, and scalable instruction. However, the research landscape on AI in ELL remains fragmented, multidisciplinary, and rapidly evolving, making it difficult for scholars, educators, and policymakers to grasp its full scope. To address this issue, this study applies bibliometric and LDA modelling analysis as a systematic and data-driven method to map the development and structure of AI-related research in ELL. The contribution of this study is a comprehensive overview of publication trends, research themes, disciplinary intersections, and influential contributors in AI and ELL, offering a foundation for future investigations and policy decisions. The study utilized data from the Scopus database covering publications from 2015 to 2025, yielding 1,510 documents. Bibliometric techniques including citation analysis, keyword co-occurrence, and ISCED-based disciplinary classification were applied using RStudio and the Bibliometrix package. Topic modeling was conducted using Latent Dirichlet Allocation (LDA) to identify thematic clusters in the literature. Findings indicate an annual growth rate of 60.47% in the volume of publications, with China, India, and Malaysia leading as primary contributors. Citation analysis indicates that key papers from 2018 and 2023 continue to shape the field's development. Using LDA (Latent Dirichlet Allocation), seven thematic clusters were detected: adaptive instruction, AI-driven writing feedback, motivation within learners, and application of machine learning. Disciplinary mapping illustrates that the foundations for AI in ELL come from Education and Humanities, along with increasing input from Computer Science and Engineering, marking its interdisciplinary global integration. This study may help curriculum designers, tech developers, and education policymakers better understand where the field is heading and what areas still need attention. It also underscores the importance of addressing challenges such as ethical considerations, data privacy, and equitable access in the use of AI for language education.