Certifying and Removing Disparate Impact

Sorelle A. Friedler, John Moeller, Carlos Scheidegger, Suresh Venkatasubramanian

2015-08-07otherACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD)1,827 citationsclosed

No readable copy stored yet — this record is metadata only.

Ethics and Social Impacts of AI Discrimination and Equality Law Artificial Intelligence in Law

Abstract

What does it mean for an algorithm to be biased? In U.S. law, unintentional bias is encoded via disparate impact, which occurs when a selection process has widely different outcomes for different groups, even as it appears to be neutral. This legal determination hinges on a definition of a protected class (ethnicity, gender) and an explicit description of the process.

Citations in this library

Stored only when both papers are in the corpus — 1,827 citations exist upstream.

References

2012Fairness through awareness

2007Uniform Guidelines on Employee Selection Procedures

Cited by

2021A Survey on Bias and Fairness in Machine Learning

2017Semantics derived automatically from language corpora contain human-like biases

2017European Union Regulations on Algorithmic Decision Making and a “Right to Explanation”

2019Fairness and Abstraction in Sociotechnical Systems

2016Man is to Computer Programmer as Woman is to Homemaker? Debiasing Word Embeddings

2019Machine behaviour

2017Algorithmic Decision Making and the Cost of Fairness

2018Fairness in Criminal Justice Risk Assessments: The State of the Art