A Learned Simulation Environment to Model Student Engagement and Retention in Automated Online Courses
This work addresses the challenge of personalizing automated online education for students, though it appears incremental as it builds on existing methods like dynamic matrix factorization.
The researchers tackled the problem of optimizing exercise ordering in online courses to improve student engagement and retention, resulting in a simulator that predicts these outcomes based on exercise sequences.
We developed a simulator to quantify the effect of exercise ordering on both student engagement and retention. Our approach combines the construction of neural network representations for users and exercises using a dynamic matrix factorization method. We further created a machine learning models of success and dropout prediction. As a result, our system is able to predict student engagement and retention based on a given sequence of exercises selected. This opens the door to the development of versatile reinforcement learning agents which can substitute the role of private tutoring in exam preparation.