Publication

Abstract : This study investigates the impact of teachers' emotions on student sentiment in MOOCs and exploring the potential relationship between emotional expressions conveyed by instructors and the emotional experiences of learners. The study uses the bidirectional encoder representations from transformers (BERT) model for sentiment analysis and the convolutional neural network (CNN) model for facial emotion recognition to analyse the emotions of teachers using instructional videos, discussion forums, and written communication. The findings demonstrate a significant correlation between teachers' emotions and student sentiment in MOOCs. Instructors displaying positive emotions elicited higher levels of student engagement and satisfaction. Conversely, negative emotional expressions were associated with decreased student sentiment and reduced motivation. The study makes a significant contribution to the emerging area of emotional dynamics in e-learning by providing vital insights for developing instructional strategies, course design and development of emotionally responsive educational technologies.