About the Paper

This paper evaluates three open-source fairness toolkits against synthetic resume-screening datasets, comparing disparate-impact detection rates, computational cost, and interpretability of their outputs for non-technical stakeholders.

Keywords