We offer a graphical interpretation of unfairness in a dataset as the
presence of an unfair causal path in the causal Bayesian network representing
the data-generation mechanism. We use this viewpoint to revisit the recent
debate surrounding the COMPAS pretrial risk assessment tool and, more
generally, to point out that fairness evaluation on a model requires careful
considerations on the patterns of unfairness underlying the training data. We
show that causal Bayesian networks provide us with a powerful tool to measure
unfairness in a dataset and to design fair models in complex unfairness
scenarios.
%0 Generic
%1 chiappa2019causal
%A Chiappa, Silvia
%A Isaac, William S.
%D 2019
%K causality
%R 10.1007/978-3-030-16744-8_1
%T A Causal Bayesian Networks Viewpoint on Fairness
%U http://arxiv.org/abs/1907.06430
%X We offer a graphical interpretation of unfairness in a dataset as the
presence of an unfair causal path in the causal Bayesian network representing
the data-generation mechanism. We use this viewpoint to revisit the recent
debate surrounding the COMPAS pretrial risk assessment tool and, more
generally, to point out that fairness evaluation on a model requires careful
considerations on the patterns of unfairness underlying the training data. We
show that causal Bayesian networks provide us with a powerful tool to measure
unfairness in a dataset and to design fair models in complex unfairness
scenarios.
@misc{chiappa2019causal,
abstract = {We offer a graphical interpretation of unfairness in a dataset as the
presence of an unfair causal path in the causal Bayesian network representing
the data-generation mechanism. We use this viewpoint to revisit the recent
debate surrounding the COMPAS pretrial risk assessment tool and, more
generally, to point out that fairness evaluation on a model requires careful
considerations on the patterns of unfairness underlying the training data. We
show that causal Bayesian networks provide us with a powerful tool to measure
unfairness in a dataset and to design fair models in complex unfairness
scenarios.},
added-at = {2019-10-09T13:03:47.000+0200},
author = {Chiappa, Silvia and Isaac, William S.},
biburl = {https://www.bibsonomy.org/bibtex/230542c2093d7bbf01f2a664539160403/stuart10},
description = {A Causal Bayesian Networks Viewpoint on Fairness},
doi = {10.1007/978-3-030-16744-8_1},
interhash = {fd3888f60059ef38cd31bcb29d9c79ea},
intrahash = {30542c2093d7bbf01f2a664539160403},
keywords = {causality},
note = {cite arxiv:1907.06430},
timestamp = {2019-10-09T13:03:47.000+0200},
title = {A Causal Bayesian Networks Viewpoint on Fairness},
url = {http://arxiv.org/abs/1907.06430},
year = 2019
}