EXPLAINABLE ARTIFICIAL INTELLIGENCE IN EDUCATION: TRANSFORMING TEACHING AND LEARNING - A REVIEW

Abstract

Explainable Artificial Intelligence (XAI) has emerged as a critical area of focus in education, where transparency and trust in AI-driven decisions are essential. This systematic review analyzes the role of XAI in predicting and enhancing student performance in educational settings. A total of 102 peer-reviewed studies published between 2015 and 2025 were examined to evaluate the use of machine learning models and explainability techniques in student performance prediction. The review highlights the most commonly applied XAI methods, such as LIME and SHAP, and assesses their effectiveness in improving interpretability without significantly compromising accuracy. Key findings indicate that integrating XAI into educational systems fosters trust among stakeholders, enables more informed decision-making, and supports early identification of at-risk students. Challenges associated with model complexity, data quality, and ethical considerations are also discussed. This review provides comprehensive insights into current XAI applications in education and identifies future research opportunities aimed at developing transparent and equitable AI-based student support systems.