Abstract:Graph neural networks (GNNs) have achieved considerable success in dealing with graph-structured data by the message-passing mechanism. Actually, this mechanism relies on a fundamental assumption that the graph structure along which information propagates is perfect. However, the real-world graphs are inevitably incomplete or noisy, which violates the assumption, thus resulting in limited performance. Therefore, optimizing graph structure for GNNs is indispensable and important. Although current semi-supervised graph structure learning (GSL) methods have achieved a promising performance, the potential of labels and prior graph structure has not been fully exploited yet. Inspired by this, we examine GSL with dual reinforcement of label and prior structure in this article. Specifically, to enhance label utilization, we first propose to construct the prior label-constrained matrices to refine the graph structure by identifying label consistency. Second, to adequately leverage the prior structure to guide GSL, we develop spectral contrastive learning that extracts global properties embedded in the prior graph structure. Moreover, contrastive fusion with prior spatial structure is further adopted, which promotes the learned structure to integrate local spatial information from the prior graph. To extensively evaluate our proposal, we perform sufficient experiments on seven benchmark datasets, where experimental results confirm the effectiveness of our method and the rationality of the learned structure from various aspects.

Large Margin Graph Construction For Semi-Supervised Learning

Graph Quality Judgement: A Large Margin Expedition.

Interactive Graph Construction for Graph-Based Semi-Supervised Learning.

Learning safe graph construction from multiple graphs

Self-supervised Learning on Graphs: Contrastive, Generative,or Predictive

Semi-Supervised Learning with Close-Form Label Propagation Using a Bipartite Graph

Semi-supervised Learning via Bipartite Graph Construction with Adaptive Neighbors

Semi-supervised Classification Via Low Rank Graph

Efficient And Robust Semi-Supervised Learning Over A Sparse-Regularized Graph

Label Information Guided Graph Construction for Semi-Supervised Learning

Graph-based Semi-supervised Learning: A Comprehensive Review

Semi-supervised Learning for Convolutional Neural Networks via Online Graph Construction

Large-Scale Graph-Based Semi-Supervised Learning Via Tree Laplacian Solver

Distributionally Robust Semi-Supervised Learning Over Graphs

A New Graph Constructor for Semi-supervised Discriminant Analysis Via Group Sparsity.

Graph-BAS3Net: Boundary-Aware Semi-Supervised Segmentation Network with Bilateral Graph Convolution

Semi-Supervised Graph Structure Learning via Dual Reinforcement of Label and Prior Structure

Semi-supervised label distribution learning via projection graph embedding

Active Model Selection for Graph-Based Semi-Supervised Learning

Robust Semi-supervised Representation Learning for Graph-Structured Data.

From Cluster Assumption to Graph Convolution: Graph-based Semi-Supervised Learning Revisited