Themis: Automatically Testing Software for Discrimination
by Rico Angell, Brittany Johnson, Yuriy Brun, Alexandra Meliou
Abstract:
Bias in decisions made by modern software is becoming a common and serious problem. We present Themis, an automated test suite generator to measure two types of discrimination, including causal relationships between sensitive inputs and program behavior. We explain how Themis can measure discrimination and aid its debugging, describe a set of optimizations Themis uses to reduce test suite size, and demonstrate Themis' effectiveness on open-source software. Themis is open-source and all our evaluation data are available at http://fairness.cs.umass.edu/. See a video of Themis in action: https://youtu.be/brB8wkaUesY
Citation:
Rico Angell, Brittany Johnson, Yuriy Brun, and Alexandra Meliou, Themis: Automatically Testing Software for Discrimination, in Proceedings of the Demonstrations Track at the 26th ACM Joint European Software Engineering Conference and Symposium on the Foundations of Software Engineering (ESEC/FSE), 2018, pp. 871–875.
Bibtex:
@inproceedings{Angell18demo-fse,
  author = {Rico Angell and Brittany Johnson and Yuriy Brun and Alexandra Meliou},
  title =
  {\href{http://people.cs.umass.edu/brun/pubs/pubs/Angell18demo-fse.pdf}{Themis: Automatically Testing Software for Discrimination}}, 
  booktitle = {Proceedings of the Demonstrations Track at the 26th ACM
  Joint European Software Engineering Conference and Symposium on the
  Foundations of Software Engineering (ESEC/FSE)},
  venue = {ESEC/FSE Demo},
  address = {Lake Buena Vista, FL, USA},
  month = {November},
  date = {6--9},
  year = {2018},
  pages = {871--875},
  accept = {$\frac{16}{36} \approx 44\%$},

  doi = {10.1145/3236024.3264590},  
  note = {\href{https://doi.org/10.1145/3236024.3264590}{DOI: 10.1145/3236024.3264590}},

  abstract = {Bias in decisions made by modern software is becoming a common
  and serious problem. We present Themis, an automated test suite generator
  to measure two types of discrimination, including causal relationships
  between sensitive inputs and program behavior. We explain how Themis can
  measure discrimination and aid its debugging, describe a set of
  optimizations Themis uses to reduce test suite size, and demonstrate
  Themis' effectiveness on open-source software. Themis is open-source and
  all our evaluation data are available at http://fairness.cs.umass.edu/. See
  a video of Themis in action: https://youtu.be/brB8wkaUesY},

  fundedBy = {NSF CCF-1453474, NSF IIS-1453543, NSF CNS-1744471, NSF CCF-1763423},
}