50 Years of Test (Un)fairness: Lessons for Machine Learning

2018/11/25 by Ben Hutchinson, Margaret Mitchell · 2 voices · 3 citations
Computer Science · Social Sciences · #Digital Economy and Work Transformation #Ethics and Social Impacts of AI #cs.AI #cs.LG

paper · pdf · doi:10.1145/3287560.3287600

FAT* '19: Conference on Fairness, Accountability, and Transparency (FAT* '19), January 29--31, 2019, Atlanta, GA, USA

arxiv published 2018/11/25 · arxiv created 2018/12/03 · arxiv updated 2018/12/05 · openalex publication_date 2019/01/09 · openalex created_date 2022/07/30 · openalex updated_date 2026/07/29

Abstract

Quantitative definitions of what is unfair and what is fair have been introduced in multiple disciplines for well over 50 years, including in education, hiring, and machine learning. We trace how the notion of fairness has been defined within the testing communities of education and hiring over the past half century, exploring the cultural and social context in which different fairness definitions have emerged. In some cases, earlier definitions of fairness are similar or identical to definitions of fairness in current machine learning research, and foreshadow current formal work. In other cases, insights into what fairness means and how to measure it have largely gone overlooked. We compare past and current notions of fairness along several dimensions, including the fairness criteria, the focus of the criteria (e.g., a test, a model, or its use), the relationship of fairness to individuals, groups, and subgroups, and the mathematical method for measuring fairness (e.g., classification, regression). This work points the way towards future research and measurement of (un)fairness that builds from our modern understanding of fairness while incorporating insights from the past.

Citations

• Equality of Opportunity in Supervised Learning
• Fairness in Criminal Justice Risk Assessments: The State of the Art
• Data Decisions and Theoretical Implications when Adversarially Learning Fair Representations
• From Parity to Preference-based Notions of Fairness in Classification
• Counterfactual Fairness
• Preventing Fairness Gerrymandering: Auditing and Learning for Subgroup Fairness
• Gender Recognition or Gender Reductionism?
• InclusiveFaceNet: Improving Face Attribute Detection with Race and Gender Diversity

Cited by

• MBFair: a model-based verification methodology for detecting violations of individual fairness
• A Survey on Bias and Fairness in Machine Learning
• Towards Fairer Datasets: Filtering and Balancing the Distribution of the People Subtree in the ImageNet Hierarchy

Discussions

• 50 Years of Test (Un)Fairness: Lessons for Machine Learning [hn, 2 points, 0 comments]
• Yeah. There was a great tech paper presented at the 2019 conference on Fairness and Accountability that gave a great historical overview of the golden age of fairness measuring, and its recession in t [bsky, 1 points, 1 comments]