Fairness-Aware Federated Graph Learning
Addressing fairness and bias in federated learning settings over graph-structured data.
Venue: EAAMO 2025
Federated learning on graphs introduces unique fairness challenges due to heterogeneous client distributions and structural biases. This project develops methods to enforce fairness constraints across decentralized clients without requiring data sharing.
Key contributions:
- Fairness-aware aggregation strategies for federated graph neural networks
- Theoretical analysis of fairness-utility tradeoffs in federated settings
- Experiments on real-world social and citation networks