Explainable Artificial Intelligence for Understanding Patterns of Educational Inequality

Educational inequality remains a persistent challenge that affects learning outcomes and access to quality education worldwide. This study employs explainable artificial intelligence (XAI) to examine the patterns and determinants of academic inequality by analyzing a large dataset of student performance, curriculum types, and parental education backgrounds. Using machine learning models such as Random Forest and XGBoost, combined with SHAP (SHapley Additive Explanations) analysis, the research identifies the most influential socio-academic factors that shape student achievement. The findings reveal that both systemic and familial variables significantly affect educational outcomes: students enrolled in structured, resource-intensive curricula tend to achieve higher and more consistent performance, while those from less standardized systems exhibit greater variability. Parental education, particularly maternal educational attainment, also emerges as a strong predictor of student success, reflecting the influence of socio-economic background on learning equity. By applying XAI techniques, the study enhances model interpretability and provides transparent, data-driven insights into the mechanisms of educational disparity. These results demonstrate that explainable AI can bridge the gap between algorithmic precision and social understanding, offering a practical framework for policymakers and educators to design evidence-based strategies aimed at promoting fairness, inclusivity, and accountability within educational systems.