Multi-Label Emotion Classification of Online Learners' Reviews Using Machine Learning Text-Based Multi-Label Classification Approach

Text- based emotion recognition is one of research areas widely developed in applied computing, but it is highly limited when dealing with online learners. In this study, we evaluate the performances of 13 multi-label classification machine learning-based methods for automatic recognizing of online learners' emotions, 12 of them are problem transformation methods and 1 is an adaptation algorithm method. The experiments are carried out using a dataset of online learners' reviews sourced from Coursera and manually multi-labeled with the emotions: Enjoyment, Excitement, Satisfaction, Frustration, Boredom, and Confusion. Our best results in term of Hamming Loss and Micro-averaged F1 Score are obtained using Random Forest classifier and classifier chains approach, while the best Macro-averaged F1 Score was obtained using Decision Tree classifier and binary relevance approach.