ARTIFICIAL INTELLIGENCE-DRIVEN ADAPTIVE E-LEARNING SYSTEMS FOR PERSONALIZED STUDENT ENGAGEMENT IN DIGITAL EDUCATION PLATFORMS

Abstract

Adaptive e-learning systems increasingly use artificial intelligence, yet discussion often treats personalization as a proxy for student engagement. Existing accounts remain fragmented across learner modeling, platform
constraints, and engagement language, which encourages overbroad claims. This paper proposes a conceptual framework that links learner data signals, learner models, categories of adaptive action, engagement dimensions,
and contextual guardrails such as privacy, fairness, and transparency. The framework is illustrated using literature-backed platform patterns to show how different adaptations target distinct engagement components rather than simple activity metrics. The contribution is a disciplined organizing logic that clarifies definitions, separates adaptive design from effectiveness claims, and supports interpretable engagement reasoning without new data. The framework is intended for AI in education researchers, digital learning designers, and higher-education practitioners working in digital education platforms.