Understanding Community Bias Amplification in Graph Representation Learning
This addresses bias amplification in graph learning for fairness applications, though it is incremental as it builds on existing contrastive learning methods.
The paper tackles the problem of community bias amplification in graph representation learning, where performance disparities between classes are exacerbated, and proposes a random graph coarsening method that mitigates this bias, as demonstrated through extensive experiments on various datasets.
In this work, we discover a phenomenon of community bias amplification in graph representation learning, which refers to the exacerbation of performance bias between different classes by graph representation learning. We conduct an in-depth theoretical study of this phenomenon from a novel spectral perspective. Our analysis suggests that structural bias between communities results in varying local convergence speeds for node embeddings. This phenomenon leads to bias amplification in the classification results of downstream tasks. Based on the theoretical insights, we propose random graph coarsening, which is proved to be effective in dealing with the above issue. Finally, we propose a novel graph contrastive learning model called Random Graph Coarsening Contrastive Learning (RGCCL), which utilizes random coarsening as data augmentation and mitigates community bias by contrasting the coarsened graph with the original graph. Extensive experiments on various datasets demonstrate the advantage of our method when dealing with community bias amplification.