Mixed Graph Contrastive Network for Semi-Supervised Node Classification

An official source code for paper Mixed Graph Contrastive Network for Semi-Supervised Node Classification. Any communications or issues are welcomed. Please contact [email protected]. If you find this repository useful to your research or work, it is really appreciate to star this repository. ❤️

---

Overview

Illustration of Mixed Graph Contrastive Network for Semi-Supervised Node Classification (MGCN) mechanism.

Requirements

The proposed MGCN is implemented with python 3.8.8 on a NVIDIA 1080 Ti GPU.

Python package information is summarized in requirements.txt:

• torch==1.8.0
• tqdm==4.61.2
• numpy==1.21.0
• tensorboard==2.8.0

Quick Start

python train.py 

Citation

If you use code or datasets in this repository for your research, please cite our paper.

@article{MGCN,
  title={Mixed Graph Contrastive Network for Semi-Supervised Node Classification},
  author={Yang, Xihong and Liu, Yue and Zhou, Sihang and Liu, Xinwang and Zhu, En},
  journal={arXiv preprint arXiv:2206.02796},
  year={2022}
}