SHARP

Mitigating Degree Bias Adaptively with Hard-to-Learn Nodes in Graph Contrastive Learning

Beta Version

Replacing the standard loss (e.g. triplet loss) in graph contrastive learning with HAR_LOSS.py will implement SHARP's key idea.

Citation

@article{hu2025mitigating,
  title={Mitigating Degree Bias Adaptively with Hard-to-Learn Nodes in Graph Contrastive Learning},
  author={Hu, Jingyu and Bo, Hongbo and Hong, Jun and Liu, Xiaowei and Liu, Weiru},
  journal={arXiv preprint arXiv:2506.05214},
  year={2025}
}