Skip to content

A curated and categorized paper list of gnn-based complex graph learning.

Notifications You must be signed in to change notification settings

jingweio/complex-graph-learning-papers

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

18 Commits
 
 

Repository files navigation

GNN-based Complex Graph Learning Papers

Table of Content

Pioneer Works

Year Title Venue Paper Code
2020 Geom-GCN: Geometric Graph Convolutional Networks ICLR 2020 Link Link
2020 Beyond Homophily in Graph Neural Networks: Current Limitations and Effective Designs NeurIPS 2020 Link Link

Problem Rethinking

Year Title Venue Paper Code
2022 Is Homophily a necessity for graph neural networks? ICLR 2022 Link Link
2021 Is Heterophily A Real Nightmare For Graph Neural Networks To Do Node Classification? Arxiv 2021 Link Link
2021 On Local Aggregation in Heterophilic Graphs Arxiv 2021 Link Link

Parwise Node Similarity and Dissimilarity

Year Title Venue Paper Code
2021 Beyond Low-frequency Information in Graph Convolutional Networks AAAI 2021 Link Link
2021 Node Similarity Preserving Graph Convolutional Networks WSDM 2021 Link Link
2022 Polarized Graph Neural Networks WWW 2022 Link Link
2022 GBK-GNN: Gated Bi-Kernel Graph Neural Networks for Modeling Both Homophily and Heterophily WWW 2022 Link Link

Topology-Structure vs Node-Features

Year Title Venue Paper Code
2021 Graph Structure Estimation Neural Networks WWW 2021 Link Link
2021 WGCN: Graph Convolutional Networks with Weighted Structural Features SIGIR 2021 Link Link
2021 Graph Neural Networks with Heterophily AAAI 2021 Link Link
2022 Graph Neural Networks Beyond Compromise Between Attributeand Topology WWW 2022 Link Link
2022 ES-GNN: Generalizing Graph Neural Networks Beyond Homophily with Edge Splitting Arxiv 2022 Link Link

Spectral Filtering

Year Title Venue Paper Code
2021 Adaptive Universal Generalized PageRank Graph Neural Network ICLR 2021 Link Link
2021 BernNet: Learning Arbitrary Graph Spectral Filters via Bernstein Approximation NeurIPS 2021 Link Link
2022 Simplified Graph Convolution with Heterophily NeuraIPS 2022 Link Link
2022 How Powerful are Spectral Graph Neural Networks ICML 2022 Link Link

Node-wise Learning

Year Title Venue Paper Code
2021 Breaking the Limit of Graph Neural Networks by Improving the Assortativity of Graphs with Local Mixing Patterns KDD 2021 Link Link
2022 Meta-Weight Graph Neural Network: Push the Limits Beyond Global Homophily WWW 2022 Link Link
2022 PA-GNN: Parameter-Adaptive Graph Neural Networks ICML WS 2022 Link Link

Connection to Robustness and Oversmoothing

Year Title Venue Paper Code
2022 Neural Sheaf Diffusion: A Topological Perspective on Heterophily and Oversmoothing in GNNs NeuraIPS 2022 Link Link
2022 How does Heterophily Impact the Robustness of Graph NeuralNetworks? Theoretical Connections and Practical Implications KDD 2022 Link Link
2022 Two Sides of the Same Coin: Heterophily and Oversmoothing in Graph Convolutional Neural Networks ICDM 2022 Link Link
2022 Understanding and Improving Graph Injection Attack by Promoting Unnoticeability ICLR 2022 Link Link

Generative Unsupervised Learning

Year Title Venue Paper Code
2022 Graph Auto-Encoder Via Neighborhood Wasserstein Reconstruction ICLR 2022 Link Link
2022 Decoupled Self-supervised Learning for Non-Homophilous Graphs NeuraIPS 2022 Link Link

Benchmarks

Year Title Venue Paper Code
2021 New Benchmarks for Learning on Non-Homophilous Graphs WWW 2021 Link Link
2021 Large Scale Learning on Non-Homophilous Graphs: New Benchmarks and Strong Simple Methods NeurIPS 2021 Link Link

Others Categories

Year Title Venue Paper Code
2021 Universal Graph Convolutional Networks NeurIPS 2021 Link Link
2021 Effective Eigendecomposition based Graph Adaptation for Heterophilic Networks Arxiv 2021 Link Link
2021 Label-informed Graph Structure Learning for Node Classification CIKM 2021 Link Link
2022 Block Modeling-Guided Graph Convolutional Neural Networks AAAI 2022 Link Link
2022 Deformable Graph Convolutional Networks AAAI 2022 Link Link
2022 Powerful Graph Convolutioal Networks with Adaptive Propagation Mechanism for Homophily and Heterophily AAAI 2022 Link Link
2022 Label-Wise Message Passing Graph Neural Network on Heterophilic Graphs Arxiv 2022 Link Link

Web Resource

[Blog] Revisiting the problem of heterophily for GNNs