AI-Driven Personalized Learning and Remedial Recommendation Through Knowledge Concept-Centric Evaluation

Abstract

This paper presents an AI-driven framework for personalized learning and remedial recommendation grounded in knowledge concept-centric evaluation. The proposed framework comprises four key modules: Knowledge Gap Identification, Adaptive Question Generation, Personalized Evaluation, and Remedial Recommendation, each designed to address individual learner needs dynamically. Retrieval-Augmented Generation (RAG) techniques are employed to create contextually aligned multiple-choice questions, while Knowledge Tracing (EKT) models continuously monitor learner progress across concepts. Experimental results based on real-world classroom deployments demonstrate significant improvements in learner competency levels following personalized interventions. Furthermore, the system supports Outcome-Based Education (OBE) principles by aligning assessments with course outcomes and learning objectives. By promoting individualized support and continuous learning, the framework advances the objectives of Sustainable Development Goal 4 (SDG 4), fostering inclusive and equitable quality education. This work highlights the potential of AI to enhance educational assessment, close learning gaps, and support scalable, competency-focused learning environments.