From Data to Insight: Decoding Student Cognition with Explainable AI Models

Abstract

Understanding student cognition is critical for enhancing educational outcomes and creating adaptive learning environments. This paper presents a comprehensive study on the application of Explainable Artificial Intelligence (XAI) models in decoding student cognitive abilities using educational data. By integrating explainable machine learning techniques with educational data mining (EDM), the proposed framework enables accurate assessment of students’ learning behaviors while maintaining transparency and interpretability. The study employs models such as SHAP (Shapley Additive Explanations), LIME (Local Interpretable Model-Agnostic Explanations), and attention-based neural networks to analyze diverse cognitive indicators including attention span, problem-solving skills, and knowledge retention. Experimental results on real-world educational datasets show significant improvements in both prediction accuracy and interpretability compared to traditional black-box models. The findings highlight the potential of XAI to provide educators with actionable insights, enabling personalized learning interventions while ensuring fairness, accountability, and trust in educational AI systems.