Interpretable Knowledge Tracing via Response Influence-based Counterfactual Reasoning

This paper proposes RCKT, a counterfactual reasoning framework for knowledge tracing that measures how individual student responses influence predictions of future performance, improving both accuracy and interpretability of student knowledge state modeling in intelligent tutoring systems.

Knowledge tracing (KT) plays a crucial role in computer-aided education and intelligent tutoring systems, aiming to assess students' knowledge proficiency by predicting their future performance on new questions based on their past response records. While existing deep learning knowledge tracing (DLKT) methods have significantly improved prediction accuracy and achieved state-of-the-art results, they often suffer from a lack of interpretability. To address this limitation, current approaches have

Source

Framework Categories

Tool Types

Tags