Artificial intelligence, Big Data and education / Kheira Ouassif

Public ISBD Titre : Artificial intelligence, Big Data and education : designing intelligent tools for analysis and guidance Type de document : document multimédia Auteurs : Kheira Ouassif, Auteur ; Benameur Ziani, Directeur de thèse Editeur : Laghouat : Université Amar Telidji - Département d'informatique Année de publication : 2025 Importance : 121 p. Accompagnement : 1 disque optique numérique (CD-ROM) Note générale : Option : Computer networks and systems Langues : Anglais Mots-clés : Educational data mining (EDM)  Machine learning  Predictive modeling  Deep neural networks (DNNs)  Apriori algorithm  Recommendation systems Résumé : In the face of growing challenges in higher education—such as rising dropout rates, academic underperformance, and the misalignment between students’ abilities and their academic paths—there is an urgent need for intelligent, data-driven solutions. This thesis explores the integration of Artificial Intelligence (AI) and Big Data analytics in the educational domain, with a focus on designing and implementing intelligent tools to support academic analysis and personalized guidance. Specifically, the research addresses critical issues within the Algerian context, where students often face limited support in making informed educational choices. The work proposes a hybrid analytical framework combining association rule mining (Apriori algorithm) with Deep Neural Networks (DNNs) to predict academic performance and guide university major selection. In the first phase, interpretable rules are extracted to identify significant academic, behavioral, and socio-economic factors influencing student outcomes. These insights are then used to enhance the learning and predictive capabilities of the DNN model, which achieves high accuracy in forecasting university GPA and optimal major placement. Complementing this predictive approach, the second phase employs clustering techniques to analyze student profiles, uncovering patterns in learner engagement and behavior. These profiles inform a recommender system designed to provide adaptive academic strategies tailored to individual learners. The proposed system was validated through experimental studies, demonstrating improved precision in student classification and increased relevance in personalized recommendations. The key contributions of this research include the development of a context-aware, hybrid predictive model for educational guidance; the integration of interpretable and deep learning techniques; and a data-informed recommendation system grounded in real-world student data. This thesis advances the field of Educational Data Mining by offering a scalable, actionable, and ethically conscious framework that supports students, educators, and institutions in making better educational decisions. The findings have practical implications for academic advising, curriculum design, and policy-making, particularly in developing educational systems seeking to embrace AI-driven innovation. note de thèses : Thèse de doctorat en informatique Artificial intelligence, Big Data and education : designing intelligent tools for analysis and guidance [document multimédia] / Kheira Ouassif, Auteur ; Benameur Ziani, Directeur de thèse . - Laghouat : Université Amar Telidji - Département d'informatique, 2025 . - 121 p. + 1 disque optique numérique (CD-ROM). Option : Computer networks and systems Langues : Anglais Mots-clés : Educational data mining (EDM)  Machine learning  Predictive modeling  Deep neural networks (DNNs)  Apriori algorithm  Recommendation systems Résumé : In the face of growing challenges in higher education—such as rising dropout rates, academic underperformance, and the misalignment between students’ abilities and their academic paths—there is an urgent need for intelligent, data-driven solutions. This thesis explores the integration of Artificial Intelligence (AI) and Big Data analytics in the educational domain, with a focus on designing and implementing intelligent tools to support academic analysis and personalized guidance. Specifically, the research addresses critical issues within the Algerian context, where students often face limited support in making informed educational choices. The work proposes a hybrid analytical framework combining association rule mining (Apriori algorithm) with Deep Neural Networks (DNNs) to predict academic performance and guide university major selection. In the first phase, interpretable rules are extracted to identify significant academic, behavioral, and socio-economic factors influencing student outcomes. These insights are then used to enhance the learning and predictive capabilities of the DNN model, which achieves high accuracy in forecasting university GPA and optimal major placement. Complementing this predictive approach, the second phase employs clustering techniques to analyze student profiles, uncovering patterns in learner engagement and behavior. These profiles inform a recommender system designed to provide adaptive academic strategies tailored to individual learners. The proposed system was validated through experimental studies, demonstrating improved precision in student classification and increased relevance in personalized recommendations. The key contributions of this research include the development of a context-aware, hybrid predictive model for educational guidance; the integration of interpretable and deep learning techniques; and a data-informed recommendation system grounded in real-world student data. This thesis advances the field of Educational Data Mining by offering a scalable, actionable, and ethically conscious framework that supports students, educators, and institutions in making better educational decisions. The findings have practical implications for academic advising, curriculum design, and policy-making, particularly in developing educational systems seeking to embrace AI-driven innovation. note de thèses : Thèse de doctorat en informatique

---

page 1/1