HyperGCL: Multi-Modal Graph Contrastive Learning via Learnable Hypergraph Views

2025-02-18

AI Summary

Existing graph contrastive learning (GCL) relies on predefined augmentations (e.g., node dropping, edge perturbation), which often impair task-relevant information, and employs coarse-grained, non-adaptive negative sample construction. To address these limitations, we propose the first learnable hypergraph-based multimodal contrastive learning framework for GCL. Our method: (1) constructs three learnable hypergraph views by jointly modeling structural and attribute information, enabling semantic-aware multimodal representation; (2) introduces an adaptive topological augmentation mechanism that dynamically preserves discriminative substructures; and (3) designs a network-aware contrastive loss that dynamically weights positive and negative sample pairs based on node neighborhood similarity. Extensive experiments on multiple benchmark datasets demonstrate state-of-the-art performance on node classification, significantly outperforming mainstream GCL approaches.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Data Mining & Knowledge Management: Graph Mining, Social Network Analysis & Community
• Computer Vision: Multi-modal Vision

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Recent advancements in Graph Contrastive Learning (GCL) have demonstrated remarkable effectiveness in improving graph representations. However, relying on predefined augmentations (e.g., node dropping, edge perturbation, attribute masking) may result in the loss of task-relevant information and a lack of adaptability to diverse input data. Furthermore, the selection of negative samples remains rarely explored. In this paper, we introduce HyperGCL, a novel multimodal GCL framework from a hypergraph perspective. HyperGCL constructs three distinct hypergraph views by jointly utilizing the input graph's structure and attributes, enabling a comprehensive integration of multiple modalities in contrastive learning. A learnable adaptive topology augmentation technique enhances these views by preserving important relations and filtering out noise. View-specific encoders capture essential characteristics from each view, while a network-aware contrastive loss leverages the underlying topology to define positive and negative samples effectively. Extensive experiments on benchmark datasets demonstrate that HyperGCL achieves state-of-the-art node classification performance.

Problem

• Improves graph representations via learnable hypergraph views
• Preserves task-relevant information in graph contrastive learning
• Enhances adaptability to diverse input data in GCL

Innovation

• Learnable hypergraph views
• Adaptive topology augmentation
• Network-aware contrastive loss