TY - GEN
T1 - Equipping Federated Graph Neural Networks with Structure-aware Group Fairness
AU - Cui, Nan
AU - Wang, Xiuling
AU - Wang, Wendy Hui
AU - Chen, Violet
AU - Ning, Yue
N1 - Publisher Copyright:
© 2023 IEEE.
PY - 2023
Y1 - 2023
N2 - Graph Neural Networks (GNNs) are used for graph data processing across various domains. Centralized training of GNNs often faces challenges due to privacy and regulatory issues, making federated learning (FL) a preferred solution in a distributed paradigm. However, GNNs may inherit biases from training data, causing these biases to propagate to the global model in distributed scenarios. To address this issue, we introduce F2GNN, a Fair Federated Graph Neural Network, to enhance group fairness. Recognizing that bias originates from both data and algorithms, F2GNN aims to mitigate both types of bias under federated settings. We offer theoretical insights into the relationship between data bias and statistical fairness metrics in GNNs. Building on our theoretical analysis, F2GNN features a fairness-aware local model update scheme and a fairness-weighted global model update scheme, considering both data bias and local model fairness during aggregation. Empirical evaluations show F2GNN outperforms SOTA baselines in fairness and accuracy.
AB - Graph Neural Networks (GNNs) are used for graph data processing across various domains. Centralized training of GNNs often faces challenges due to privacy and regulatory issues, making federated learning (FL) a preferred solution in a distributed paradigm. However, GNNs may inherit biases from training data, causing these biases to propagate to the global model in distributed scenarios. To address this issue, we introduce F2GNN, a Fair Federated Graph Neural Network, to enhance group fairness. Recognizing that bias originates from both data and algorithms, F2GNN aims to mitigate both types of bias under federated settings. We offer theoretical insights into the relationship between data bias and statistical fairness metrics in GNNs. Building on our theoretical analysis, F2GNN features a fairness-aware local model update scheme and a fairness-weighted global model update scheme, considering both data bias and local model fairness during aggregation. Empirical evaluations show F2GNN outperforms SOTA baselines in fairness and accuracy.
KW - Federated Learning
KW - Graph Neural Networks
KW - Group Fairness
UR - http://www.scopus.com/inward/record.url?scp=85185396531&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85185396531&partnerID=8YFLogxK
U2 - 10.1109/ICDM58522.2023.00111
DO - 10.1109/ICDM58522.2023.00111
M3 - Conference contribution
AN - SCOPUS:85185396531
T3 - Proceedings - IEEE International Conference on Data Mining, ICDM
SP - 980
EP - 985
BT - Proceedings - 23rd IEEE International Conference on Data Mining, ICDM 2023
A2 - Chen, Guihai
A2 - Khan, Latifur
A2 - Gao, Xiaofeng
A2 - Qiu, Meikang
A2 - Pedrycz, Witold
A2 - Wu, Xindong
T2 - 23rd IEEE International Conference on Data Mining, ICDM 2023
Y2 - 1 December 2023 through 4 December 2023
ER -