A Review of Emotion Recognition in Virtual Learning Environments and its Educational Impact

Abstract

Understanding and addressing students' emotional needs is crucial in the rapidly evolving domain of online learning, as it fosters students' motivation, interest, and educational outcomes. This literature review examines the methods, findings, and implications of recent studies that attempt to identify and analyze emotions in online learning contexts. Methodologically, a systematic review approach was employed to analyze a wide variety of academic publications released between 2021 and 2024. The survey-encompassing studies employed various methods to recognize emotions, such as happiness, sadness, and interest in virtual learning environments, including physiological signal analysis, deep learning models, and machine learning algorithms. The outcome of the literature review points out significant progress in the area of emotion detection technology where studies depict how effectively deep learning and machine learning models can recognize and interpret students' emotional expression along with effectively identifying them, Finding from the reviewed papers shows that models like CNN, LSTM, SVM, ViT, and brain-computer interfaces have been employed with varying degrees of accuracy (ranging from 55% to over 90%). In addition, using real-time feedback mechanisms that recognize emotions has the potential to improve learning outcomes, motivation, and student engagement in online learning environments