ELU-GCN: Effectively Label-Utilizing Graph Convolutional Network

2024-11-04

arXiv.org

AI Summary

Graph Convolutional Networks (GCNs) suffer from insufficient utilization of label information and weak label influence on unlabeled nodes. Method: This paper proposes a two-stage framework: (1) a label-driven graph structure self-adaptation mechanism to construct an Enhanced Label-propagation Utility (ELU) graph, improving label diffusion capability; and (2) dual-graph contrastive learning between the original graph and the ELU graph, incorporating cross-graph message passing and consistency constraints to enhance representation robustness. Contribution/Results: Theoretical analysis demonstrates that the proposed method improves the generalization bound of GCNs. Extensive experiments on multiple benchmark datasets show significant performance gains over state-of-the-art GCN variants and graph contrastive learning methods, validating the synergistic improvement in label utilization efficiency and node representation capability.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Search and Optimization: Learning to Search
• Computer Vision: Representation Learning for 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
• Economics, Online Markets and Human Computation: Humans versus LLMs for data annotation and labeling

Abstract

The message-passing mechanism of graph convolutional networks (i.e., GCNs) enables label information to be propagated to a broader range of neighbors, thereby increasing the utilization of labels. However, the label information is not always effectively utilized in the traditional GCN framework. To address this issue, we propose a new two-step framework called ELU-GCN. In the first stage, ELU-GCN conducts graph learning to learn a new graph structure (ie ELU-graph), which enables GCNs to effectively utilize label information. In the second stage, we design a new graph contrastive learning on the GCN framework for representation learning by exploring the consistency and mutually exclusive information between the learned ELU graph and the original graph. Moreover, we theoretically demonstrate that the proposed method can ensure the generalization ability of GCNs. Extensive experiments validate the superiority of the proposed method.

Problem

• Improving label propagation in GCNs for unlabeled nodes
• Enhancing label utilization via new graph structure learning
• Ensuring GCN generalization with contrastive learning

Innovation

• Proposes ELU-GCN for better label utilization
• Learns new graph structure (ELU-graph) first
• Uses graph contrastive learning for representation