Enhancing Online Teaching and Learning Through Machine Learning and Learning Analytics

Abstract

Online learning environments are gaining popularity as an alternative to traditional learning environments during times of global pandemic. The traditional classroom setting has given way to online learning in recent educational developments. To employ it as a classifier in online education, the goal of this work is to construct a real-time face emotion detection system that recognizes and categorizes emotions of a human. It uses machine learning to identify, forecast, and analyse a learner's facial expressions, and it further maps those expressions to a learning affect that categorizes the emotions of those who are seen on camera. Academic feelings can have a significant impact on learning outcomes. Students typically show their emotions through their facial expressions, voice, and behavior. A spontaneous facial expression database is created in light of the inference algorithm's lack of training samples. It consists of two subsets: a video clip database and a picture database, and it includes the typical emotional facial expressions. The database contains 1,274 video clips and 30,184 photos from 82 pupils. The parameters for student facial detection can be used to gauge each learner's rate of concentration. The performances of the SVM and RF classifier in facial expression for the image are presented. The results of testing this with the RF algorithm and a 90.14% accuracy rate produced extremely good results.