Enhancing Personalized Learning with Frequent Pattern Mining of Educational Data

Le résumé fourni par la source

Personalized learning systems aim to adapt educational content, pace, and delivery to individual student needs, addressing diverse learning preferences and academic backgrounds. However, current models rely on complex fuzzy or hybrid frameworks that hinder scalability and interpretability, particularly in real-time learning environments. This paper presents a lightweight, data-driven approach using Frequent Pattern Mining (FPM), specifically the FP-Growth and Apriori algorithm, to enhance personalization. Leveraging the Open University Learning Analytics Dataset (OULAD), this work proposes a framework supporting four core tasks: identifying common learning behaviors, recommending personalized content, predicting subject-specific challenges, and enabling adaptive learning paths. The experimental results show that frequent pattern-based rules achieve up to 77 % accuracy in predicting student engagement and 74 % in analyzing student behavior. Compared to complex models, the proposed method provides a scalable, interpretable, and efficient alternative for building responsive and student-centered learning systems.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.