Data Management for Causal Algorithmic Fairness

Abstract

Fairness is increasingly recognized as a critical component of machine learning systems. However, it is the underlying data on which these systems are trained that often reflects discrimination, suggesting a data management problem. In this paper, we first make a distinction between associational and causal definitions of fairness in the literature and argue that the concept of fairness requires causal reasoning. We then review existing works and identify future opportunities for applying data management techniques to causal algorithmic fairness.

arXiv commentarXiv admin note: text overlap with arXiv:1902.08283

References in corpus (8)

• Equality of Opportunity in Supervised Learning
• Counterfactual Fairness
• Snorkel: Rapid Training Data Creation with Weak Supervision
• Fairness Testing: Testing Software for Discrimination
• A survey on measuring indirect discrimination in machine learning
• Causal Reasoning for Algorithmic Fairness
• Quantifying Causal Effects on Query Answering in Databases
• Capuchin: Causal Database Repair for Algorithmic Fairness