Online Student Assessment Based on Facial Expression Recognition through Machine Learning

Engagement is a major aspect of smart teaching since it reveals how much students are learning. Compared to traditional face-to-face classrooms, online courses lack exact, quick, and relevant communication and feedback between lecturers and students. Existing data structures for automated student interaction are geared at analyzing participation, which is crucial to the interface's ability to provide appropriate feedback. Students may struggle with data structures and algorithms due to their novelty, yet the topic is important because of the strategic model teaching framework that includes automated teaching of data structures via group-based collaborative interactions among peers. This paper suggests the Novel Facial Expression Recognition Model (NFERM) based on Machine Learning to analyze student engagement and evaluate student assessment in online education. Students are “engaged” when they maintain high attention, curiosity, enthusiasm, optimism, and passion for what they are taught or learning. Evaluating students' progress in an eLearning course is best done via activities including quizzes, matching tests, self-assessments, studies, questions for issue resolution, questions based on scenarios, and games. To better identify interactions in photos, this research offers a machine learning approach that addresses the issue of data imbalance by first training on publicly accessible facial expression data. This research presents a virtual computer's-eye-view of a proposed architecture for incorporating facial expression recognition (FER) algorithm into online learning platforms.