Interpreting Unfairness in Graph Neural Networks via Training Node Attribution

Authors

  • Yushun Dong University of Virginia
  • Song Wang University of Virginia
  • Jing Ma University of Virginia
  • Ninghao Liu University of Georgia
  • Jundong Li University of Virginia

DOI:

https://doi.org/10.1609/aaai.v37i6.25905

Keywords:

ML: Graph-based Machine Learning, PEAI: Bias, Fairness & Equity

Abstract

Graph Neural Networks (GNNs) have emerged as the leading paradigm for solving graph analytical problems in various real-world applications. Nevertheless, GNNs could potentially render biased predictions towards certain demographic subgroups. Understanding how the bias in predictions arises is critical, as it guides the design of GNN debiasing mechanisms. However, most existing works overwhelmingly focus on GNN debiasing, but fall short on explaining how such bias is induced. In this paper, we study a novel problem of interpreting GNN unfairness through attributing it to the influence of training nodes. Specifically, we propose a novel strategy named Probabilistic Distribution Disparity (PDD) to measure the bias exhibited in GNNs, and develop an algorithm to efficiently estimate the influence of each training node on such bias. We verify the validity of PDD and the effectiveness of influence estimation through experiments on real-world datasets. Finally, we also demonstrate how the proposed framework could be used for debiasing GNNs. Open-source code can be found at https://github.com/yushundong/BIND.

Downloads

Published

2023-06-26

How to Cite

Dong, Y., Wang, S., Ma, J., Liu, N., & Li, J. (2023). Interpreting Unfairness in Graph Neural Networks via Training Node Attribution. Proceedings of the AAAI Conference on Artificial Intelligence, 37(6), 7441-7449. https://doi.org/10.1609/aaai.v37i6.25905

Issue

Section

AAAI Technical Track on Machine Learning I