Improvement of Neural Cognitive Diagnosis for Intelligent Education Systems Based on Response Time Factor

Cognitive diagnosis is a fundamental issue in intelligent education, which aims to discover the proficiency level of students on specific knowledge concepts. In this area, the Neural Cognitive Diagnosis (NeuralCD) framework is a good solution. However, the NeuralCD model does not consider the response time of the students when answering questions. In this paper, we proposed an improved neural cognitive diagnostic model by introducing a response time factor into the NeuralCD model, named NeuralCDM-RTF. The core of the response time factor is the answering duration of the student on a problem. The duration of the response time showed the student’s proficiency indirectly. We proposed the concept of response time factor and proposed trapezoid model to calculate the response time factor. By performing experiments on real-world datasets, the results show that the NeuralCDM-RTF has higher accuracy than the traditional NeuralCD model while the improved model remains interpretable.