Self-attention based GRU neural network for deep knowledge tracing
Shangzhu Jin, Yan Zhao, Jun Peng, Ning Chen, Run Xue, Minghui Liang, Yunfeng Jiang · 2022 IEEE 17th Conference on Industrial Electronics and Applications (ICIEA) · 2022
How to provide students with personalized guidance and personalized teaching is a key concern in the field of education. Knowledge tracing can analyze students' exercise records, trace students' learning situations in real-time, obtain students' knowledge mastery, predict students' future performance, and provide a new way to teach students in accordance with their aptitude. Aiming at the problem that the traditional model loses information under long sequence data, this paper uses GRU neural network combined with attention mechanism to realize deep knowledge tracing model, which improves the sensitivity of the model to long sequence data, strengthens the influence of historical knowledge state on future answer performance, and improves the prediction accuracy of knowledge mastery. The experiment has achieved good results in public datasets.