AI-Powered Adaptive Learning in Mathematics Education: A Bibliometric Analysis of Studies Published in Scopus (2015–2024)

Artificial intelligence (AI)–driven adaptive learning optimises mathematics instruction by calibrating curricular items, pacing, and evaluative prompts in real-time according to the ongoing learner state. This bibliometric study delineates the trajectory of Scopus-indexed scholarship from 2015 to 2024, applying a rigorously designed search string, a venue-filtering procedure to minimise cross-disciplinary spillover, and a stepwise data-certification protocol. Analysis encompasses temporal trends, journal leadership, geographic author contributions, and emergent keyword co-occurrence. A novel conceptual framework is forwarded to position adaptivity within a formative-assessment paradigm. Results reveal a post-2020 growth inflection and a paradigm migration from heuristic-rule systems to learning-embedded policies. Uniform figure and caption formatting, sequential numbering, and APA-7 citation throughout address prior editorial critiques; a bounded limitations discussion and a dedicated ethics declaration align with submission norms.