Exercise recommendation based on knowledge concept prediction

Good recommendation for difficulty exercises can effectively help to point the students/users in the right direction, and potentially empower their learning interests. It is however challenging to select the exercises with reasonable difficulty for students as they have different learning status and the size of exercise bank is quite large. The classic collaborative filtering (CF) based recommendation methods rely heavily on the similarities among students or exercises, leading to recommend exercises with mismatched difficulty. This paper proposes a novel exercise recommendation method, which uses Recurrent Neural Networks (RNNs) to predict the coverage of knowledge concepts, and uses Deep Knowledge Tracing (DKT) to predict students' mastery level of knowledge concepts based on the student's exercise answer records. The predictive results are utilized to filter the exercises; therefore, a subset of exercise bank is generated. As such, a complete list of recommended exercises can be obtained by solving an optimization problem. Extensive experimental studies show that our proposed approach has advantages over some existing baseline methods, not only in terms of the evaluation of difficulty of recommended exercises, but also the diversity and novelty of the recommendation lists.