Student Next Assignment Submission Prediction Using a Machine Learning Approach

Abstract

The web-based learning platform provides quality education nowadays, but assignment submission is a critical issue in the e-learning system. Therefore, to investigate assignment submission of the student in advance before the end of course is an important problem. The assignment submission prediction is the advantage of the e-learning system because it allows the instructor to find students’ problems on time. Additionally, online learning mostly depends on demographic characteristics such as region, age, and education level. This study uses machine learning (ML) methods to detect students who do not submit assignments on time and then also find which demographic factors affects online assignment submission. The data is publicly available and was collected from an open university of U.K. The result shows that Random Forest is an optimal option for predicting students who do not submit an assignment on time. We have also found that Gender, Student Credit, Final result, Total clicks, Score are strong predictors for student assignment.