Kecerdasan Artifisial Generatif dalam Pembelajaran: Analisis Bibliometrik Tren Penelitian Global 2023-2026

Abstract

This bibliometric study analyzes global research trends in generative artificial intelligence (AI) for education using 27 Scopus-indexed articles from 2023-2026. Data analysis was conducted in R Studio using bibliometrix package following PRISMA-ScR protocol. Results show a negative annual growth rate of -20.63%, reflecting stringent selection criteria rather than declining research interest. The average of 6.45 authors per document indicates strong interdisciplinary collaboration, while international collaboration remains limited at 21.05%. China dominates publication output, followed by Australia and USA. Four thematic clusters were identified: generative AI in learning, student self-efficacy, engineering education, and educational technology acceptance. Strategic thematic mapping reveals "students," "artificial intelligence," and "higher education" as motor themes, while adversarial machine learning and contrastive learning emerge as niche themes. The study concludes that fragmented research networks require enhanced international collaboration and deeper exploration of emerging themes. These findings provide a strategic roadmap for researchers, policymakers, and educators to address knowledge gaps and prioritize future research agendas in AI-enhanced education.