<![CDATA[The rapid digital transformation of education has accelerated the use of artificial intelligence, particularly deep learning, to support adaptive and personalized learning systems. However, research in this field remains fragmented, and its overall development has not been systematically mapped. This study aims to identify trends, gaps, and future directions of deep learning in education between 2000 and 2025. A scoping review was conducted on 491 journal articles indexed in Scopus, followed by bibliometric analysis to examine publication growth, keyword co-occurrence, and thematic clusters. The results show a sharp increase in publications since 2018, reflecting growing academic attention to this field. Four dominant clusters were identified: (1) academic evaluation and assessment, (2) learner perceptions and experiences, (3) algorithmic exploration and model efficiency, and (4) reflective and systemic approaches to technology-based learning. While technical and evaluative studies dominate, ethical, affective, and social dimensions remain underexplored, particularly in primary education and Global South contexts. The findings highlight the need for future research to adopt a more transdisciplinary and contextualized approach, integrating pedagogy, digital ethics, and human-centered design. This study contributes strategically by providing a comprehensive knowledge map to guide researchers, policymakers, and practitioners in designing equitable and sustainable AI-based education systems.]]>
