Reproduce “Hidden Disparities in Ranking: Category-Aware Fairness via Distributional Stability” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Gayane Taturyan, Charlotte Laclau, Stephan Clémençon
The authors present a method that mixes existing ranking policies to reduce group utility gaps within item categories while targeting category-level service. It measures utility using exposure to relevant items and optimizes policy weights under changing group shares. In a synthetic job-platform simulation, conditional utility parity (CUP) keeps category gaps near zero; this illustrates the method, not real-world job disparities.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.