Détection Automatique de l’Engagement Émotionnel dans un Environnement d’Apprentissage Humain en Utilisant les Techniques du Machine Learning

Abstract

Emotions play a central role in human interaction and understanding, and their accurate detection has a significant impact in fields such as education, mental health, and user in- terfaces. Studies have shown that emotional engagement is crucial in improving learning outcomes in online learning platforms. In this work, we propose an approach for the automatic detection of learners’ emotional engagement in online learning platforms. The proposed approach is based on analyz- ing learners’ feedback to detect their emotions. Support Vector Machines (SVM), Naive Bayes classification (NB), and the Bidirectional Encoder Representations from Transform- ers (BERT) model have been used in emotion detection. The BERT model has shown great capability in detecting emotions from natural language.