<p>Recent research in learning analytics has increasingly focused on identifying “at-risk” students to enable timely and meaningful interventions. While predictive analytics has played a central role in this effort, the integration of explainable artificial intelligence (xAI) has emerged more recently as a way to enhance the transparency and interpretability of predictive models. This paper proposes a framework that integrates clustering, classification, and xAI techniques to support the generation of transparent and personalized recommendations aimed at enhancing student learning progress. The approach consists of four sequential phases: (i) clustering (ii) classifying (iii) generation and evaluation of explainable rules, and (iv) intervention generation based on the associated rules. A key contribution of the proposed framework is its ability to provide transparency and learner-friendly generated insights. To achieve this, a rule-based system was developed to transform xAI-based rules into clear, and human-readable feedback for each student. The proposed framework was empirically validated using real, longitudinal learner data, examining learner achievement patterns over time, and visualizing temporal dynamics. This validation shows that interpretable analytics can provide deep insights into student learning progression and offer practical value for improving teaching strategies and learning outcomes.</p>