This repository contains the resources on graph neural network (GNN) considering heterophily.

Boost learning for GNNs from the graph structure under challenging heterophily settings. (NeurIPS'20)

A Survey of Learning from Graphs with Heterophily

Dir-GNN is a machine learning model that enables learning on directed graphs.

NeurIPS 2022, Revisiting Heterophily For Graph Neural Networks, official PyTorch implementation for Adaptive Channel Mixing (ACM) GNN framework

Gradient gating (ICLR 2023)

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

Code for GBK-GNN (paper accepted by WWW2022)

How does Heterophily Impact the Robustness of Graph Neural Networks? Theoretical Connections and Practical Implications (KDD'22)

Implementation of GCNH, a GNN for heterophilous graphs described in the paper "GCNH: A Simple Method For Representation Learning On Heterophilous Graphs", IJCNN 2023

Homophily on Human Interaction via. Stochastic Block Model network representing the likelihood of gathering of homophily people contributing to human connection with common likeliness by social status.

Neighbourhood standardization as a means to improve GNN performance heterophilic datasets.