DAGKT: Difficulty and Attempts Boosted Graph-based Knowledge Tracing

TL;DR

This paper introduces DAGKT, a novel graph-based knowledge tracing model that incorporates question difficulty and student attempts, leveraging graph neural networks and question similarity to improve student mastery prediction.

Contribution

DAGKT is the first to integrate question difficulty and attempts into graph neural network-based knowledge tracing, enhancing predictive accuracy using rich student record data.

Findings

DAGKT outperforms existing KT models on three real-world datasets.

Incorporating difficulty and attempts improves model accuracy.

Question similarity based on F1 score enhances graph construction.

Abstract

In the field of intelligent education, knowledge tracing (KT) has attracted increasing attention, which estimates and traces students' mastery of knowledge concepts to provide high-quality education. In KT, there are natural graph structures among questions and knowledge concepts so some studies explored the application of graph neural networks (GNNs) to improve the performance of the KT models which have not used graph structure. However, most of them ignored both the questions' difficulties and students' attempts at questions. Actually, questions with the same knowledge concepts have different difficulties, and students' different attempts also represent different knowledge mastery. In this paper, we propose a difficulty and attempts boosted graph-based KT (DAGKT), using rich information from students' records. Moreover, a novel method is designed to establish the question similarity relationship inspired by the F1 score. Extensive experiments on three real-world datasets demonstrate the effectiveness of the proposed DAGKT.