Leveraging Peer-Informed Label Consistency for Robust Graph Neural Networks with Noisy Labels

Li, Kailai; Sun, Jiawei; Lou, Jiong; Feng, Zhanbo; Zhou, Hefeng; Wu, Chentao; Xue, Guangtao; Zhao, Wei; Li, Jie

doi:10.24963/ijcai.2025/623

Leveraging Peer-Informed Label Consistency for Robust Graph Neural Networks with Noisy Labels

Kailai Li, Jiawei Sun, Jiong Lou, Zhanbo Feng, Hefeng Zhou, Chentao Wu, Guangtao Xue, Wei Zhao, Jie Li

Download BibTeX Cite

Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence. Main Track. Pages 5598-5606.

https://doi.org/10.24963/ijcai.2025/623

Keywords

Machine Learning: ML: Sequence and graph learning
Machine Learning: ML: Weakly supervised learning
Machine Learning: ML: Robustness
Machine Learning: ML: Other

Abstract

Graph Neural Networks (GNNs) excel in many applications but struggle when trained with noisy labels, especially as noise can propagate through the graph structure.

Despite recent progress in developing robust GNNs, few methods exploit the intrinsic properties of graph data to filter out noise.

In this paper, we introduce ProCon, a novel framework that identifies mislabeled nodes by measuring label consistency among semantically similar peers, which are determined by feature similarity and graph adjacency.

Mislabeled nodes typically exhibit lower consistency with these peers, a signal we measure using pseudo-labels derived from representational prototypes.

A Gaussian Mixture Model is fitted to the consistency distribution to identify clean samples, which refine prototype quality in an iterative feedback loop.

Experiments on multiple datasets demonstrate that ProCon significantly outperforms state-of-the-art methods, effectively mitigating label noise and enhancing GNN robustness.

Kailai Li, Jiawei Sun, Jiong Lou, Zhanbo Feng, Hefeng Zhou, Chentao Wu, Guangtao Xue, Wei Zhao, and Jie Li. 2025. Leveraging Peer-Informed Label Consistency for Robust Graph Neural Networks with Noisy Labels. In Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence, IJCAI-25. IJCAI, 5598–5606. https://doi.org/10.24963/ijcai.2025/623

@inproceedings{ijcai2025p623,
  title     = {Leveraging Peer-Informed Label Consistency for Robust Graph Neural Networks with Noisy Labels},
  author    = {Li, Kailai and Sun, Jiawei and Lou, Jiong and Feng, Zhanbo and Zhou, Hefeng and Wu, Chentao and Xue, Guangtao and Zhao, Wei and Li, Jie},
  booktitle = {Proceedings of the Thirty-Fourth International Joint Conference on
               Artificial Intelligence, {IJCAI-25}},
  publisher = {International Joint Conferences on Artificial Intelligence Organization},
  editor    = {James Kwok},
  pages     = {5598--5606},
  year      = {2025},
  month     = {8},
  note      = {Main Track},
  doi       = {10.24963/ijcai.2025/623},
  url       = {https://doi.org/10.24963/ijcai.2025/623},
}

TY  - CPAPER
TI  - Leveraging Peer-Informed Label Consistency for Robust Graph Neural Networks with Noisy Labels
AU  - Li, Kailai
AU  - Sun, Jiawei
AU  - Lou, Jiong
AU  - Feng, Zhanbo
AU  - Zhou, Hefeng
AU  - Wu, Chentao
AU  - Xue, Guangtao
AU  - Zhao, Wei
AU  - Li, Jie
T2  - Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence, IJCAI-25
PB  - International Joint Conferences on Artificial Intelligence Organization
PY  - 2025
SP  - 5598
EP  - 5606
DO  - 10.24963/ijcai.2025/623
UR  - https://doi.org/10.24963/ijcai.2025/623
ER  -