tengxiao1/awesome-graph-contrastive-learning-with-heterophily

Papers about Graph Contrastive Learning and Graph Self-supervised Learning on Graphs with Heterophily

Awesome-Graph Contrastive Learning-with-Heterophily

This repository contains a list of the relevant resources on Graph Contrastive Learning and Graph Self-supervised Learning on graphs with heterophily. We categorize the following papers based on their published years. We will try our best to continuously maintain this repository in real time.

If you found any error or any missed paper, please don't hesitate to open issues.

Graphs with Heterophily (Heterophilous Graphs or Non-Homophilous Graphs)

Homophily is a key principle of many real-world graphs, whereby linked nodes often belong to the same class or have similar features (“birds of a feather flock together”). For example, friends are likely to have similar political beliefs, and papers tend to cite papers from the same research area. However, in the real world, there are also settings where “opposites attract”, leading to graphs with heterophily: linked nodes are likely from different classes or have dissimilar features. For instance, the majority of people tend to connect with people of the opposite gender in dating graphs, different amino acid types are more likely to connect in proteins, fraudsters are more likely to connect to accomplices than to other fraudsters in purchasing networks.

--Zhu et al [NeurIPS 2020].

Graph Contrastive Learning and Graph Self-supervised Learning on unlabled heterophilous graphs is an emerging research area.

Year 2023

1. [NeurIPS 2023] Simple and Asymmetric Graph Contrastive Learning without Augmentations [paper] [code]
2. [CIKM 2023] MUSE: Multi-View Contrastive Learning for Heterophilic Graphs [paper] [code]
3. [AAAI 2023] Beyond Smoothing: Unsupervised Graph Representation Learning with Edge Heterophily Discriminating [paper] [code]
4. [Arxiv 2023] Contrastive Learning under Heterophily [paper]
5. [ICML 2023] Contrastive Learning Meets Homophily: Two Birds with One Stone [paper]
6. [TKDD 2022] Revisiting the Role of Heterophily in Graph Representation Learning: An Edge Classification Perspective [paper]
7. [TKDE 2023] Graph Representation Learning Beyond Node and Homophily [paper]

Year 2022

1. [NeurIPS 2022] Decoupled Self-supervised Learning for Non-Homophilous Graphs [paper] [code]
2. [Arxiv 2022] Can Single-Pass Contrastive Learning Work for Both Homophilic and Heterophilic Graph? [paper]
3. [CIKM 2022] Towards Self-supervised Learning on Graphs with Heterophily [paper] [code]
4. [ICLR 2022] Graph Auto-Encoder Via Neighborhood Wasserstein Reconstruction [paper] [code]