Context-aware nonlinear and neural attentive knowledge-based models for grade prediction

Abstract

Grade prediction can help students and their advisers select courses and design personalized degree programs based on predicted future course performance. One of the successful approaches for accu- rately predicting a student's grades in future courses is Cumulative Knowledge-based Regression Models (CKRM). CKRM learns shallow linear models that predict a student's grades as the similarity between his/her knowledge state and the target course. However, there can be more complex interactions among prior courses taken by a student, which cannot be captured by the current linear CKRM model. More- over, CKRM and other grade prediction methods ignore the effect of concurrently-taken courses on a student's performance in a target course. In this paper, we propose context-aware nonlinear and neural attentive models that can potentially better estimate a student's knowledge state from his/her prior course information, as well as model the interactions between a target course and concurrent courses. Compared to the competing methods, our experiments on a large real-world dataset consisting of more than 1.5 million grades show the effectiveness of the proposed models in accurately predicting students' grades. Moreover, the attention weights learned by the neural attentive model can be helpful in better designing their degree plans.