This paper defines the notions of software fairness and discrimination and develops a testing-based method for measuring if and how much software discriminates. Specifically, the paper focuses on measuring causality in discriminatory behavior. Modern software contributes to important societal decisions and evidence of software discrimination has been found in systems that recommend criminal sentences, grant access to financial loans and products, and determine who is allowed to participate in promotions and receive services. Our approach, Themis, measures discrimination in software by generating efficient, discrimination-testing test suites. Given a schema describing valid system inputs, Themis generates discrimination tests automatically and, notably, does not require an oracle. We evaluate Themis on 20 software systems, 12 of which come from prior work with explicit focus on avoiding discrimination. We find that (1) Themis is effective at discovering software discrimination, (2) state-of-the-art techniques for removing discrimination from algorithms fail in many situations, at times discriminating against as much as 98% of an input subdomain, (3) Themis optimizations are effective at producing efficient test suites for measuring discrimination, and (4) Themis is more efficient on systems that exhibit more discrimination. We thus demonstrate that fairness testing is a critical aspect of the software development cycle in domains with possible discrimination and provide initial tools for measuring software discrimination.
@inproceedings{Galhotra17fse,
author = {Sainyam Galhotra and Yuriy Brun and Alexandra Meliou},
title = {\href{http://people.cs.umass.edu/brun/pubs/pubs/Galhotra17fse.pdf}{Fairness Testing: Testing Software for Discrimination}},
booktitle = {Proceedings of the 11th Joint Meeting of the European
Software Engineering Conference and ACM SIGSOFT Symposium on the
Foundations of Software Engineering (ESEC/FSE)},
venue = {ESEC/FSE},
month = {September},
year = {2017},
date = {6--8},
address = {Paderborn, Germany},
pages = {498--510},
doi = {10.1145/3106237.3106277},
note = {\raisebox{-.5ex}{\includegraphics[height=2.5ex]{trophy}}~ACM SIGSOFT Distinguished Paper Award.
\href{https://doi.org/10.1145/3106237.3106277}{DOI: 10.1145/3106237.3106277},
arXiv: \href{https://arxiv.org/abs/1709.03221}{abs/1709.03221}},
comment = {<span class="emphasis">ACM SIGSOFT Distinguished Paper Award</span>},
accept = {$\frac{72}{295} \approx 24\%$},
abstract = {<p>This paper defines the notions of software fairness and
discrimination and develops a testing-based method for measuring if and how
much software discriminates. Specifically, the paper focuses on measuring
causality in discriminatory behavior. Modern software contributes to
important societal decisions and evidence of software discrimination has
been found in systems that recommend criminal sentences, grant access to
financial loans and products, and determine who is allowed to participate
in promotions and receive services. Our approach, Themis, measures
discrimination in software by generating efficient, discrimination-testing
test suites. Given a schema describing valid system inputs, Themis
generates discrimination tests automatically and, notably, does not require
an oracle. We evaluate Themis on 20 software systems, 12 of which come from
prior work with explicit focus on avoiding discrimination. We find that
(1) Themis is effective at discovering software discrimination,
(2) state-of-the-art techniques for removing discrimination from algorithms
fail in many situations, at times discriminating against as much as 98% of
an input subdomain, (3) Themis optimizations are effective at producing
efficient test suites for measuring discrimination, and (4) Themis is more
efficient on systems that exhibit more discrimination. We thus demonstrate
that fairness testing is a critical aspect of the software development
cycle in domains with possible discrimination and provide initial tools for
measuring software discrimination.</p>},
fundedBy = {NSF CCF-1453474, NSF IIS-1453543, NSF CNS-1744471},
}