Measuring Epistemic Unfairness for Algorithmic Decision-Making

This document lists three recent arXiv papers. 1) “Measuring Epistemic Unfairness for Algorithmic Decision-Making” proposes a quantitative framework to measure epistemic injustices in algorithmic systems (credibility, uptake, epistemic agency). The authors map deficits to stages of algorithmic mediation, adapt distributive fairness indices to epistemic goods, distinguish resource vs. capability/rights stances, and demonstrate the approach in a recommender-mediated opinion-dynamics simulation—showing epistemic harms can persist even when standard predictive fairness metrics are satisfied. Imply