When Contrastive Learning Meets Active Learning: A Novel Graph Active Learning Paradigm with Self-Supervision

2020/10/30 by Yanqiao Zhu, Weizhi Xu, Zhu, Yanqiao +5 · 3 citations
Computer Science · Mathematics · #Advanced Graph Neural Networks #Artificial intelligence #Computer science #Domain Adaptation and Few-Shot Learning #Expressive power #FOS: Computer and information sciences #Feature learning #Graph #Leverage (statistics) #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Machine Learning and Algorithms #Machine learning #Theoretical computer science #cs.LG #stat.ML

paper · pdf · doi:10.48550/arxiv.2010.16091

published in arXiv (Cornell University) (Cornell University) · Preliminary work, 16 pages

openalex publication_date 2020/10/30 · arxiv created 2021/04/16 · arxiv updated 2021/04/19 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

This paper studies active learning (AL) on graphs, whose purpose is to discover the most informative nodes to maximize the performance of graph neural networks (GNNs). Previously, most graph AL methods focus on learning node representations from a carefully selected labeled dataset with large amount of unlabeled data neglected. Motivated by the success of contrastive learning (CL), we propose a novel paradigm that seamlessly integrates graph AL with CL. While being able to leverage the power of abundant unlabeled data in a self-supervised manner, nodes selected by AL further provide semantic information that can better guide representation learning. Besides, previous work measures the informativeness of nodes without considering the neighborhood propagation scheme of GNNs, so that noisy nodes may be selected. We argue that due to the smoothing nature of GNNs, the central nodes from homophilous subgraphs should benefit the model training most. To this end, we present a minimax selection scheme that explicitly harnesses neighborhood information and discover homophilous subgraphs to facilitate active selection. Comprehensive, confounding-free experiments on five public datasets demonstrate the superiority of our method over state-of-the-arts.

Citations

• Active Learning for Graph Embedding
• Learning deep representations by mutual information estimation and maximization
• Fast Graph Representation Learning with PyTorch Geometric
• Detecting Functionality-Specific Vulnerabilities via Retrieving Individual Functionality-Equivalent APIs in Open-Source Repositories
• PyTorch: An Imperative Style, High-Performance Deep Learning Library
• Active Learning for Graph Neural Networks via Node Feature Propagation
• Deep Graph Contrastive Representation Learning
• Simple and Deep Graph Convolutional Networks
• Momentum Contrast for Unsupervised Visual Representation Learning

Related

• GraphCL: Contrastive Self-Supervised Learning of Graph Representations
• Adversarial Graph Augmentation to Improve Graph Contrastive Learning
• Learning Graph Representation by Aggregating Subgraphs via Mutual Information Maximization
• Active Learning for Graphs with Noisy Structures
• Rethinking the Promotion Brought by Contrastive Learning to Semi-Supervised Node Classification