Enhancing knowledge tracing with question-based contrastive learning

Xiaoxuan Shen, Fenghua Yu, Yaqi Liu, Ruxia Liang, Qian Wan, Tianhao Yang, Mengtian Shi, Jianwen Sun

Abstract: Highlights•We propose a novel approach, QCKT, which represents questions from two different perspectives.•We propose five contrastive learning strategies to uncover similarities and differences between questions.•We extend 13 KT methods on 6 publicly available datasets. Extensive experiments demonstrate that QCKT significantly improves their performance.