Explainable Artificial Intelligence Framework for Personalised Learning Gap Identification in K–12 Education
Abstract
Artificial intelligence-based predictive models have demonstrated significant potential for identifying students at risk of academic difficulties; however, limited interpretability remains a major challenge for their practical adoption in educational environments. This study developed an explainable artificial intelligence (XAI) framework to interpret machine learning predictions and support personalised identification of learning gaps among K–12 students. The best-performing machine learning model obtained from predictive analysis was integrated with explainability techniques to determine the contribution of individual academic, behavioural, and engagement-related factors influencing learning-gap predictions. The framework incorporated feature importance analysis and SHAP (SHapley Additive exPlanations) interpretation to provide transparent explanations of model decisions at both global and student-specific levels. The results demonstrated that previous academic achievement, assessment performance trends, assignment completion behaviour, and engagement patterns were the primary contributors to learning-gap identification. Individual prediction analysis further showed that AI-generated explanations could identify specific factors responsible for increased learning risk and provide meaningful guidance for targeted educational support. By transforming complex machine learning predictions into understandable educational insights, the proposed framework enhanced transparency and trust in AI-assisted decision-making. The findings highlight the importance of explainable AI in developing responsible educational technologies where artificial intelligence supports teachers in recognising learner needs and designing personalised interventions. This research contributes to human-centred AI applications in education by presenting an interpretable framework for early learning-gap detection and evidence-based educational support.
How to Cite
Nidhi Bhatia, Dr Sarika Sharma, "Explainable Artificial Intelligence Framework for Personalised Learning Gap Identification in K–12 Education", Vol. 4, Issue 4, 25-07-2026, pp. 60-70.
