DS-HGCN: A Dual-Stream Hypergraph Convolutional Network for Predicting Student Engagement via Social Contagion

TL;DR

This work addresses the challenge of predicting student engagement by capturing social contagion within classroom groups. It introduces DS-HGCN, a dual-stream hypergraph convolutional network with a multivariate propagation path for high-order interactions and a parallel multi-frequency path for frequency-aware feature propagation, augmented by a hypergraph attention mechanism. The approach demonstrates state-of-the-art performance on the RoomReader dataset for both binary and ternary engagement classification, outperforming unimodal, multi-feature, and existing graph-based methods. The results underscore the importance of modeling group-level contagion and frequency-specific information for robust engagement prediction in educational settings.

Abstract

Student engagement is a critical factor influencing academic success and learning outcomes. Accurately predicting student engagement is essential for optimizing teaching strategies and providing personalized interventions. However, most approaches focus on single-dimensional feature analysis and assessing engagement based on individual student factors. In this work, we propose a dual-stream multi-feature fusion model based on hypergraph convolutional networks (DS-HGCN), incorporating social contagion of student engagement. DS-HGCN enables accurate prediction of student engagement states by modeling multi-dimensional features and their propagation mechanisms between students. The framework constructs a hypergraph structure to encode engagement contagion among students and captures the emotional and behavioral differences and commonalities by multi-frequency signals. Furthermore, we introduce a hypergraph attention mechanism to dynamically weigh the influence of each student, accounting for individual differences in the propagation process. Extensive experiments on public benchmark datasets demonstrate that our proposed method achieves superior performance and significantly outperforms existing state-of-the-art approaches.

Figures (6)

• Figure 1: This figure presents the schematic illustration of classroom student engagement prediction. The DS-HGCN leverages the multi-dimensional features and the engagement contagion of students to infer their engagement status.
• Figure 2: The pipeline of the proposed DS-HGCN.
• Figure 3: The illustration of the hyperedge convolution layer, which can perform node-edge-node transform, better refining the features using the hypergraph structure.
• Figure 4: The illustration of the MFGCN is presented. It aggregates low-frequency and high-frequency signals from neighboring nodes.
• Figure 5: Heatmaps illustrating the performance on binary and ternary classification tasks with varying numbers of hypergraph layers ($L$) and graph layers ($K$).