Adaptive Multi-view Graph Contrastive Learning via Fractional-order Neural Diffusion Networks

2025-11-09

AI Summary

Existing graph contrastive learning methods rely on handcrafted, fixed views (e.g., local/global), limiting their ability to adaptively capture multi-scale structural patterns. To address this, we propose an adaptive multi-view graph contrastive learning framework based on fractional-order neural diffusion networks. Our method employs a learnable fractional-order derivative α ∈ (0,1] to continuously modulate the scope of information propagation, enabling end-to-end generation of diverse node representations spanning local to global scales. By parameterizing diffusion scales as a continuous dynamical process, it eliminates dependence on discrete view design and manual data augmentation. Integrating fractional calculus, continuous-time graph neural networks, and contrastive learning, our approach significantly enhances representation discriminability and robustness. Extensive experiments demonstrate consistent and substantial improvements over state-of-the-art graph contrastive learning methods across standard benchmarks.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Computer Vision: Diffusion Models for Vision
• Search and Optimization: Learning to Search

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Graph contrastive learning (GCL) learns node and graph representations by contrasting multiple views of the same graph. Existing methods typically rely on fixed, handcrafted views-usually a local and a global perspective, which limits their ability to capture multi-scale structural patterns. We present an augmentation-free, multi-view GCL framework grounded in fractional-order continuous dynamics. By varying the fractional derivative order $alpha in (0,1]$, our encoders produce a continuous spectrum of views: small $alpha$ yields localized features, while large $alpha$ induces broader, global aggregation. We treat $alpha$ as a learnable parameter so the model can adapt diffusion scales to the data and automatically discover informative views. This principled approach generates diverse, complementary representations without manual augmentations. Extensive experiments on standard benchmarks demonstrate that our method produces more robust and expressive embeddings and outperforms state-of-the-art GCL baselines.

Problem

• Adaptively captures multi-scale structural patterns in graphs
• Learns optimal diffusion scales without manual view augmentation
• Generates robust embeddings through fractional-order neural diffusion

Innovation

• Fractional-order neural diffusion generates multi-scale graph views
• Learnable fractional derivative adapts diffusion scales automatically
• Augmentation-free contrastive learning creates diverse graph representations