Augmenting Personalized Question Recommendation with Hierarchical Information for Online Test Platform

Lu Jiang, Wanfei Zhang, Yibin Wang, Na Luo, Lin Yue

Abstract: Personalized question recommendation for students is an important research topic in the field of smart education. Current studies depend on collaborative filtering based, cognitive diagnosis based, or cognitive diagnosis based on collaborative filtering methods. However, the above methods can only model the knowledge state for a single student and the common features of similar students while ignoring students’ flat and hierarchical information. To solve the problems above, we propose an augmenting personalized question recommendation method(APQR) which combines flat and hierarchical information. Firstly, we propose a framework to capture student and question hierarchical information jointly. Secondly, we propose a cognitive diagnostic method that uses flat and hierarchical information to model students’ proficiency on each question. Finally, we recommend questions based on students’ performance by using probabilistic matrix factorization combined with students’ proficiency. We apply APQR to personalized question recommendation to demonstrate the performance improvement via an online test platform dataset. The promising results show that the proposed APQR can recommend questions to students effectively.