Bibliometric Dataset on Artificial Intelligence in Sustainable Education: Bibliometric Analysis
Description
This dataset comprises metadata of 116 scientific publications on the integration of Artificial Intelligence (AI) in sustainable education, published between 2021 and April 2026. The data were retrieved from the Scopus database using a targeted search query focusing on AI‑related terms, sustainable education terms, and social sciences and psychology subject areas. This dataset formed the basis for a bibliometric analysis examining publication trends, influential authors and journals, collaborative networks, thematic evolution, and citation impact in the field. The analysis employed Biblioshiny (R package) and VOSviewer software, following the PRISMA protocol.
Files
Steps to reproduce
Data Source and Retrieval: Bibliographic data were extracted from the Scopus database on April 13, 2026. The search query targeted publications on artificial intelligence in sustainable education within the subject areas of Social Sciences and Psychology from 2021 onwards. The query combined three keyword groups: AI terms: TITLE-ABS-KEY(("artificial intelligence" OR "AI" OR "machine learning")) Sustainable education terms: TITLE-ABS-KEY(("sustainable education" OR "education for sustainable development" OR "ESD")) The search was limited to publication years 2021–April 2026, document type (articles), language (English), and subject areas (Social Sciences and Psychology). Screening and Selection: The initial search yielded 586 documents. These were screened following PRISMA guidelines: Filtering by year (2021–April 2026) reduced the set to 506 records. Limiting to articles and English language resulted in 271 records. Applying subject area filters (Social Sciences and Psychology) left 125 records. Abstracts were manually reviewed for relevance to the intersection of AI and sustainable education. Nine records were excluded due to lack of relevance. This resulted in a final dataset of 116 articles. The dataset (in .csv or .xlsx format) comprises the bibliographic records (authors, titles, abstracts, keywords, citations, affiliations, source titles) of these 116 articles. Data Analysis: The dataset was analyzed using: Biblioshiny (R package for bibliometrix): For quantitative bibliometric analysis (publication trends, author/journal productivity, citation analysis, three‑field plots). VOSviewer (version 1.6.20): For network visualization and clustering (co‑authorship networks, keyword co‑occurrence for thematic mapping, with a minimum keyword occurrence of two). This dataset forms the basis for the study "Artificial Intelligence in Sustainable Education: A Bibliometric Analysis".