August 16, 2025Applied SciencesOpen Access

Hypergraph-Driven High-Order Knowledge Tracing with a Dual-Gated Dynamic Mechanism

View Full Paper

Ask AI

Bookmark

The HGKT model significantly enhances prediction accuracy in knowledge tracing with dynamic learning mechanisms.
Achieving a state-of-the-art AUC of 0.78372, HGKT outperforms existing models across three datasets, including EdNet-KT1.
Utilizing hypergraph convolution, HGKT captures high-order exercise-concept correlations and integrates temporal features.
Visualization analyses demonstrate HGKT's effectiveness in interpreting knowledge state evolution in educational contexts.

Discussion

No takes yet. Share an insight, caveat, or question.

Novel hypergraph-driven model improves predictive performance in knowledge tracing, highlighting dynamic learning behaviors.

Ma et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark