MTGKT: Multiscale Temporal Graph Knowledge Tracing

doi:10.1109/tlt.2026.3707679

MTGKT: Multiscale Temporal Graph Knowledge Tracing

2026-06-26

0

PRE

AI

DOI：10.1109/tlt.2026.3707679

Original

Original request for help

Save

Abstract

En 中文

Knowledge tracing (KT) dynamically monitors students’ evolving knowledge mastery states through historical interaction sequences and predicts future performance. Existing KT methods often assume homogeneous temporal effects, and therefore, fail to fully capture the heterogeneous impact of multiscale intervals on knowledge dependence evolution. To tackle this challenge, we propose the multiscale temporal graph KT (MTGKT) model. First, we design a multiscale temporal encoder that decomposes intervals into micro-, meso-, and macroscales and fuses them through soft assignment. Second, we introduce an adaptive forgetting mechanism based on the Ebbinghaus forgetting curve, with both knowledge-specific and student-specific parameters to model individualized memory decay. Third, we devise an enhanced dynamic graph neural network with structured neighbor sampling, skill cooccurrence features, and message passing, together with gated recurrent unit and multihead attention for temporal dynamics. In experiments, we conducted comparative evaluations between MTGKT and nine representative KT models on four public datasets, and the results show competitive performance under the reported experimental settings.

Keywords:

Adaptive forgetting mechanism

dynamic graph neural network (GNN)

knowledge tracing (KT)

multihead attention

multiscale time encoder

MTGKT: Multiscale Temporal Graph Knowledge Tracing

Abstract

Journal

Organization