Samia Drissi

Dropout prediction in Massive Open Online Courses (MOOCs) has been extensively studied, yet existing systems share three fundamental limitations: Accurate models are black boxes, post-hoc explanations approximate rather than faithfully represent model decisions, and predictions are rarely translated into concrete instructor actions. This paper presents NeSy-Drop, a neuro-symbolic framework that simultaneously addresses prediction, explanation, and personalized intervention routing for MOOC dropout. NeSy-Drop constructs a heterogeneous graph per course cohort encoding student–resource–assessment interactions, processed through a heterogeneous graph transformer encoder, five behavioral atom predictor MLPs, and a differentiable symbolic rule layer producing guaranteed faithful ante-hoc explanations. A three-level explainability stack provides symbolic rule chains, SHAP embedding attribution, LIME raw-feature importance, and gradient-based counterfactual prescriptions. Each at-risk student is routed to one of five concrete interventions at one of three severity levels. Evaluated on OULAD covering 32,593 students across 22 cohorts, NeSy-Drop achieves AUC of 0.961 and macro F1 of 0.8983, within 2.2% AUC of the best non-interpretable baseline under a fair evaluation protocol, while being the only system that simultaneously predicts, explains, and prescribes actions at the individual student level.

The rapid advancement of digital technologies and the widespread use of the Internet have transformed the educational landscape, positioning distance learning as a central component of modern education. However, this shift has introduced several challenges that negatively affect student motivation and academic performance. To address these issues, this paper proposes an intelligent e-learning environment that provides personalized support tailored to individual learner needs. The study focuses on adaptive assistance for students experiencing difficulties by analyzing their interactions to identify specific learning barriers. Preliminary results indicate that the proposed approach enhances interactivity and engagement, leading to significant improvements in online learners' performance.

The rise of Massive Open Online Courses (MOOCs) has democratised education, yet student dropout remains a persistent challenge. This research tackles this issue by analysing 1,273 publications on MOOC dropout rates (2020-2023) from the Scopus database using VOSviewer, a bibliometric analysis tool. By mapping the intellectual landscape, the study identifies key research questions (e.g., concerning MOOC dropout, learning motivation) and emerging themes (e.g., classification performance, engagement). This comprehensive analysis provides a roadmap for future research, pinpointing areas for further investigation and knowledge advancement.