Counterpart Fairness--Addressing Systematic between-group Differences in Fairness Evaluation
We analyzed the impact of systematic differences and biases in data collection on group fairness assessment and accordingly proposed a counterpart-based fairness evaluation index.
Nov 19, 2024
On dyadic fairness: Exploring and mitigating bias in graph connections.
We theoretically relate the graph connections to dyadic fairness on link predictive scores in learning graph neural networks and accordingly introduced an algorithm for fair link prediction by adjusting the adjacency weight matrix to address the fairness-utility trade-off.
Jan 12, 2021