AI evidence extraction

Mapping the AI Surge in Higher Education: A Bibliometric Study Spanning a Decade (2015–2025)

Mousin Omarsaib, Sara Bibi Mitha, Anisa Vahed, Ghulam Masudh Mohamed · Informatics · 2025

AI-generated evidence extraction, verified across multiple analytical personas. Not a substitute for the peer-reviewed original.

7/10

Relevance

0/4

Quality (LMQS)

Evidence

Citations

0.00

FWCI

This is an AI analysis. Read the peer-reviewed original at the publisher: https://doi.org/10.3390/informatics12040137

Methodology & findings

Study design

Bibliometric analysis combining Web of Science (WoS) and Scopus datasets synthesized through Biblioshiny platform, co-occurrence network analysis in VOSviewer, and Latent Dirichlet Allocation (LDA) for thematic extraction from locally cited references..

Sample

> 1000

Primary method

Biblioshiny platform for data synthesis, VOSviewer for co-occurrence network analysis, and Latent Dirichlet Allocation (LDA) for topic modeling and thematic extraction from locally cited references.

Main result

The study revealed that "Findings revealed significant acceleration patterns from 2023 concerning publication trends, annual growth patterns, cited references, top authors, leading journals, and leading countries." Additionally, "Patterns of strengths from co-occurrence networks in VOSviewer revealed growing interest in generative AI tools, AI ethics, and concerns about AI integration into the curriculum in HE."

Reports effect sizes.

Research paradigm

Positivist/quantitative bibliometric analysis

Author conclusions

The authors conclude that "The development of frameworks and ethical guidelines are important to address fair and transparent adoption of AI in HE. Further, global inequalities in adoption, aligning with UNESCO's Sustainable Development Goals, are crucial to ensure equitable and responsible AI integration in HE."

Risk of bias

Database selection bias (WoS and Scopus only; excludes other sources); Temporal scope bias (2015-2025 may miss earlier foundational work); Language bias (likely English-only publications in major databases); Publication bias (positive findings overrepresented in indexed journals); Database selection bias: only WoS and Scopus used, excluding other academic databases; Language bias: likely English-language publication bias in indexed databases; Publication bias: bibliometric studies only capture published literature; Temporal bias: focused on 2015-2025 period, may miss foundational work outside this window

Open questions raised

Recommendations for future work include: (1) policymakers and stakeholders addressing pedagogical integration of generative AI tools in HE; (2) development of frameworks and ethical guidelines for fair and transparent AI adoption; (3) addressing global inequalities in AI adoption aligned with UNESCO's Sustainable Development Goals to ensure equitable and responsible AI integration in HE.
The authors identify that policymakers and stakeholders need to address pedagogical integration of generative AI tools in HE, develop frameworks and ethical guidelines for fair and transparent adoption of AI, and address global inequalities in AI adoption to ensure equitable and responsible AI integration aligned with UNESCO's Sustainable Development Goals.
Precise mapping of AI growth loci in higher education remains unclear
Need for frameworks and ethical guidelines for fair AI adoption in HE
Global inequalities in AI adoption require attention aligned with UNESCO's SDGs
Equitable and responsible AI integration in HE needs further development

Data: not_statedCode: not_statedExtracted from: pdfAgreement 63%

Explore related topics

Institutional Policy AI Governance Meta-Analysis