Engagement Detection during E-Learning Classes using Machine Learning Algorithms

Abstract

Engagement detection during e-learning classes is super important because it helps us understand how students learn and stay interested in their studies. Lately, computers have been getting really good at figuring out if students are paying attention during online classes. In this research, we're using facial expressions and eye movements to see if students are focused during online classes. We're comparing different computer programs to figure out which ones are best at doing this job. We're also looking at how different features, like facial expressions and where students are looking, affect how well these programs work.

Student engagement refers to students’ psychological involvement in the acquiring and comprehending academic work tries to teach you stuff, like what you know, what you can do, and the skills you have. In realm of education, student engagement stands as a important factor influencing academic success and overall learning outcomes. Student accomplishment is directly proportionate to his or her level of engagement.  The main issue we face here is that it is not feasible for the teacher to teach in the online classroom and simultaneously notice which students are distracted and which are not. Distracted students not only fail to learn themselves but are also much more likely to cause disturbances during the class. . By using machine learning algorithms to detect engagement levels during e-learning classes, educators can identify students who may be struggling and provide them with additional support. Our study also highlights the importance of certain features for engagement detection during e-learning classes, which can inform the design of more effective e-learning environments.

Our study helps us understand how computer programs can tell if students are paying attention in online classes. This helps make online learning better in the future.

Model was developed to identify and inform inattentive learners. We must get the input data from

this video stream before doing the analysis, and the model must be as successful under dim lighting

Faculty -Jagbeer Singh, Department of CSE, MIET, Meerut.

[1] Create a mechanism to measure students levels of involvement based on their eye movements.

[2] Based on the amount of participation classify into three groups: ”extremely involved,” ”usually engaged,” and ”not engaged