Deep Learning for Dropout Prediction in MOOCs

Abstract

In recent years, the rapid rise of massive open online courses (MOOCs) has aroused great attention. Dropout prediction or identifying students at risk of dropping out of a course is an open problem for MOOC researchers and providers. This paper formulates the dropout prediction problem as predicting how much content in the whole course syllabus can be completed by the student. A dropout rate prediction model is based on a recurrent neural network (RNN), and an URL embedding layer is proposed to solve this problem. The results show that the prediction accuracy of the model is significantly higher than that of the traditional machine learning model.