Thesis defense of Nesrine MANSOURI

The rapid evolution of digital learning ecosystems has profoundly transformed the way educational content is delivered, accessed and personalized. In this context, Competency-Based Education (CBE) is emerging as a key paradigm, emphasizing measurable learning outcomes and mastery of skills, rather than mere course validation.

However, as programs become more modular and organized around blocks of competencies, students encounter difficulties in identifying optimal learning pathways that match their abilities, interests and career goals. This thesis proposes an artificial intelligence-based framework designed to recommend personalized learning paths based on skill blocks, exploiting machine learning techniques to support decision-making on both the learner and teacher sides.

The research work presented is based on three main contributions:

1. A predictive model of student performance (SMOTE + GA + GRU) was developed to identify students at risk of academic failure, enabling early and targeted pedagogical interventions.
2. A specialization recommendation system (SBS + Adaboost + GA) has been implemented to guide students towards the specialization best suited to their profile and prior learning.
3. A MOOC recommendation model (SFS + Decision Tree) was designed to help learners identify the most relevant online courses, thus addressing the problem of information overload and fostering engagement in digital learning environments.

Together, these contributions constitute a unified approach demonstrating the potential of hybrid and data-driven methods to enhance the personalization of education and support lifelong learning.

By integrating data preprocessing, feature selection and explainable machine learning techniques, this research contributes to the development of intelligent, skill-sensitive recommender systems, promoting more adaptive, transparent and learner-centered education.

Mr. MOURAD ABED Professeur des universités, Université Polytechnique Hauts de France, Thesis Director
. M. Makram SOUI Assistant professor, University of Michigan, Thesis co-supervisor
.
Mrs. Zaineb GARCIA, Professor, University of Lille, Rapporteur
Mr. Issam NOUAOURI, Professor, University of Lille, Co-supervisor M. Issam NOUAOURI, Professeur des universités, Université d'Artois, Rapporteur

Ms Hounaida SAKLY, Associate Professor, Centre de Recherche en Microélectronique et Nanotechnologie, Sousse, Examiner
Mr. Imed Riadh FARA, Professor, University of Sousse, Rapporteur Mr. Imed Riadh FARAH, Professor, Université de Manouba-Tunis, Examiner
Mr. Bruno DE LIEVRE, Associate Professor M. Bruno DE LIEVRE, Professor, Université de Mons, Examiner
M. Rene MANDIAU Mr. Rene MANDIAU Professeur des Universités, Université Polytechnique Hauts-de-France (UPHF), Examiner

E-learning, competency-based learning, Recommendation, Personalization, AI, Machine learning