Group fairness without demographics using social networks

Abstract

Group fairness is a popular approach to prevent unfavorable treatment of individuals based on sensitive attributes such as race, gender, and disability. However, the reliance of group fairness on access to discrete group information raises several limitations and concerns, especially with regard to privacy, intersectionality, and unforeseen biases. In this work, we propose a "group-free" measure of fairness that does not rely on sensitive attributes and, instead, is based on homophily in social networks, i.e., the common property that individuals sharing similar attributes are more likely to be connected. Our measure is group-free as it avoids recovering any form of group memberships and uses only pairwise similarities between individuals to define inequality in outcomes relative to the homophily structure in the network. We theoretically justify our measure by showing it is commensurate with the notion of additive decomposability in the economic inequality literature and also bound the impact of non-sensitive confounding attributes. Furthermore, we apply our measure to develop fair algorithms for classification, maximizing information access, and recommender systems. Our experimental results show that the proposed approach can reduce inequality among protected classes without knowledge of sensitive attribute labels. We conclude with a discussion of the limitations of our approach when applied in real-world settings.

References in corpus (30)

• Fast unfolding of communities in large networks
• Community structure in social and biological networks
• Community detection in graphs
• Finding community structure in very large networks
• Uncovering the overlapping community structure of complex networks in nature and society
• Assortative mixing in networks
• Finding community structure in networks using the eigenvectors of matrices
• Mixing patterns in networks
• Equality of Opportunity in Supervised Learning
• The Anatomy of the Facebook Social Graph
• Improving fairness in machine learning systems: What do industry practitioners need?
• Fairness of Exposure in Rankings
• Spectral clustering and the high-dimensional stochastic blockmodel
• Pseudo-likelihood methods for community detection in large sparse networks
• Multiresolution community detection for megascale networks by information-based replica correlations
• Measurement and Fairness
• Fairness Without Demographics in Repeated Loss Minimization
• Controlling Fairness and Bias in Dynamic Learning-to-Rank
• A Unified Approach to Quantifying Algorithmic Unfairness: Measuring Individual & Group Unfairness via Inequality Indices
• FairRec: Two-Sided Fairness for Personalized Recommendations in Two-Sided Platforms
• Fairness Under Unawareness: Assessing Disparity When Protected Class Is Unobserved
• Fairness for Unobserved Characteristics: Insights from Technological Impacts on Queer Communities
• Auditing for Discrimination in Algorithms Delivering Job Ads
• User Fairness, Item Fairness, and Diversity for Rankings in Two-Sided Markets
• Understanding Regularized Spectral Clustering via Graph Conductance
• Gaps in Information Access in Social Networks
• "What We Can't Measure, We Can't Understand": Challenges to Demographic Data Procurement in the Pursuit of Fairness
• Equalized odds postprocessing under imperfect group information
• Proxy Fairness
• "You Can't Fix What You Can't Measure": Privately Measuring Demographic Performance Disparities in Federated Learning