Artificial intelligence (AI) and machine learning (ML) are increasingly applied in education, yet their roles in educational quality management require further conceptual organization. This study examines publication trends and thematic structures in AI/ML research related to educational quality management and develops a literature-derived computational taxonomy of reported applications. Bibliometric analysis was conducted on 510 Scopus-indexed documents published from 2016 to 2026, with the 2026 records retrieved as of 2 October. The analysis examined publication patterns, source and author productivity, geographical distribution, keyword co-occurrence, and conceptual structure, followed by classification of applications according to computational functions, quality-management applications, input-data types, and decision outputs. Publication output increased substantially after 2023, reaching 156 documents in 2025, while 148 documents were indexed for 2026 at the retrieval date. Keyword co-occurrence analysis identified three broad thematic groupings associated with AI and educational systems, students and teaching, and quality-related concepts. Multiple Correspondence Analysis (MCA) showed that the first two dimensions accounted for a cumulative 65.26% of the total inertia. These dimensions represent associations among the categorical data analyzed and do not independently validate the proposed taxonomy. The resulting taxonomy organizes reported applications, including prediction, classification, natural language processing, recommendation, and generative AI, in relation to assessment, quality assurance, performance monitoring, feedback, and decision support. The framework provides a basis for further conceptual refinement and independent empirical validation but does not establish that AI/ML applications improve educational quality.
Copyrights © 2027