From Artificial Intelligence to E-Learning: A Bibliometric Overview of Language Teaching and Educational Technology Literature

Abstract

This study conducts a bibliometric analysis of research at the intersection of Arabic language teaching and educational technology to characterize the structural configuration and thematic evolution of the field. Data were retrieved from Scopus and Web of Science and screened following the PRISMA 2020 protocol, yielding a final corpus of 174 documents published between 2003 and 2026. Performance analysis, scientific-law analyses (Lotka and Bradford), keyword and co-occurrence analysis, co-authorship and co-citation network analyses, and Latent Dirichlet Allocation (LDA) topic modeling were conducted through the srlengine.com academic analysis platform. The corpus comprises 762 authors and a total of 802 citations (mean 4.61 per document; h-index 14, g-index 23, i10-index 20). Author productivity conforms to a Lotka distribution with an exponent of 3.65 — substantially above the classical value of 2.0 — indicating an early-stage field with a narrow productive core. Bradford analysis identified 11 core journals with a multiplier of 4.09, confirming high source concentration. Output is regionally concentrated in Malaysia, Indonesia, and Saudi Arabia, while China emerges as a citation outlier, producing only 24 publications but attracting 1,376 citations through integration with the broader AI-in-language-education literature. All collaboration and co-citation networks return low density values (0.028–0.063), depicting a fragmented intellectual structure. LDA analysis reveals that two AI-oriented topics jointly account for 44% of the corpus, whereas pronunciation and speech constitute only 1%, signaling a pronounced under-development of speech-technology and diglossia-aware research. The field is expanding rapidly under the post-2022 generative AI wave but remains structurally immature, geographically concentrated, and thematically uneven.