🤖 AI Summary
This work addresses the overlooked issue of fairness disparities among sensitive attribute groups in multi-hop graph structures when promoting inter-group connections for link prediction. The paper introduces the concept of k-hop fairness and, for the first time, formalizes a structural bias metric that accounts for multi-hop distances, thereby revealing the intrinsic dependence between fairness and graph topology. This approach transcends the limitations of conventional methods that focus solely on first-order neighborhoods or pairwise fairness. By designing preprocessing and postprocessing strategies based on graph rewiring, the proposed method effectively mitigates multi-hop structural bias on standard benchmarks. Experimental results demonstrate that the approach significantly outperforms existing baselines in terms of multi-hop fairness while maintaining competitive link prediction performance.
📝 Abstract
Link prediction (LP) plays a central role in graph-based applications, particularly in social recommendation. However, real-world graphs often reflect structural biases, most notably homophily, the tendency of nodes with similar attributes to connect. While this property can improve predictive performance, it also risks reinforcing existing social disparities. In response, fairness-aware LP methods have emerged, often seeking to mitigate these effects by promoting inter-group connections, that is, links between nodes with differing sensitive attributes (e.g., gender), following the principle of dyadic fairness. However, dyadic fairness overlooks potential disparities within the sensitive groups themselves. To overcome this issue, we propose $k$-hop fairness, a structural notion of fairness for LP, that assesses disparities conditioned on the distance between nodes in the graph. We formalize this notion through predictive fairness and structural bias metrics, and propose pre- and post-processing mitigation strategies. Experiments across standard LP benchmarks reveal: (1) a strong tendency of models to reproduce structural biases at different $k$-hops; (2) interdependence between structural biases at different hops when rewiring graphs; and (3) that our post-processing method achieves favorable $k$-hop performance-fairness trade-offs compared to existing fair LP baselines.