Representation Learning with Mutual Influence of Modalities for Node Classification in Multi-Modal Heterogeneous Networks

TL;DR

This paper introduces HGNN-IMA, a novel heterogeneous graph neural network that captures mutual modal influences through inter-modal attention, improving node classification in multi-modal heterogeneous networks by effectively fusing modalities and handling missing data.

Contribution

It proposes a new inter-modal attention mechanism within a heterogeneous graph transformer to enhance multi-modal fusion and node classification accuracy.

Findings

Outperforms existing methods in node classification accuracy.

Effectively handles missing modalities with augmented attention loss.

Demonstrates robustness across various multi-modal network datasets.

Abstract

Nowadays, numerous online platforms can be described as multi-modal heterogeneous networks (MMHNs), such as Douban's movie networks and Amazon's product review networks. Accurately categorizing nodes within these networks is crucial for analyzing the corresponding entities, which requires effective representation learning on nodes. However, existing multi-modal fusion methods often adopt either early fusion strategies which may lose the unique characteristics of individual modalities, or late fusion approaches overlooking the cross-modal guidance in GNN-based information propagation. In this paper, we propose a novel model for node classification in MMHNs, named Heterogeneous Graph Neural Network with Inter-Modal Attention (HGNN-IMA). It learns node representations by capturing the mutual influence of multiple modalities during the information propagation process, within the framework of heterogeneous graph transformer. Specifically, a nested inter-modal attention mechanism is integrated into the inter-node attention to achieve adaptive multi-modal fusion, and modality alignment is also taken into account to encourage the propagation among nodes with consistent similarities across all modalities. Moreover, an attention loss is augmented to mitigate the impact of missing modalities. Extensive experiments validate the superiority of the model in the node classification task, providing an innovative view to handle multi-modal data, especially when accompanied with network structures.