The DOI matches its Crossref record; 92 of the 92 references it cites were found in a registry or an index; none of the 45 sources its claims cite is checked for support yet.
Fairness testing: testing software for discrimination
Open accessAbstract
This paper defines software fairness and discrimination and develops a testing-based method for measuring if and how much software discriminates, focusing on causality in discriminatory behavior. Evidence of software discrimination has been found in modern software systems that recommend criminal sentences, grant access to financial products, and determine who is allowed to participate in promotions. Our approach, Themis, generates efficient test suites to measure discrimination. Given a schema describing valid system inputs, Themis generates discrimination tests automatically and 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.
Overview
AI snapshot
Written from the paper’s full text.
TL;DR
The paper defines causal software discrimination and presents Themis, a testing method that automatically generates test suites to measure group and causal discrimination without an oracle. Evaluating 20 instances of 8 financial-decision systems, it finds discrimination even in systems designed to avoid it and shows sound pruning reduces test suites by two-to-three orders of magnitude.
Key findings
- Themis detected causal discrimination missed by group discrimination: for `B census gender`, the causal score was 11.27% versus a group score of 0.52%, more than 21× higher.
- Discrimination was present even in systems designed to avoid it: `B census gender` had a causal discrimination score over 11% for gender, and discriminated against gender, marital status, and race with a causal score of 77.2%; `F census` discriminated against country and race with a causal score of 98.1%.
- There was no clear evidence that discrimination-aware methods outperformed discrimination-unaware ones; the discrimination-unaware approaches typically discriminated less than their discrimination-aware counterparts, with the exception of logistic regression.
- Pruning reduced the number of test cases by, on average, a factor of 2,849 for group and 148 for causal discrimination; Themis was more efficient when systems discriminated more, e.g., `B credit gender` needed 3,413 tests at threshold θ = 0.7 but only 10 tests at θ = 0.6.
Methods
- The paper formally defines group, causal, and apparent discrimination, then introduces Themis, which uses an input schema, requires no oracle, and employs test caching, adaptive confidence-driven sampling, and sound pruning based on monotonicity properties.
- Evaluation used 20 instances of 8 software systems making financial decisions: 12 instances from prior discrimination-aware work (4 focused on race, 8 on gender) and 8 standard machine-learning instances that made no attempt to avoid discrimination.
- The datasets were the Adult/Census Income dataset (45K individuals, 14 attributes) and the Statlog German Credit dataset (1,000 individuals, 20 pieces of data); systems were trained on balanced samples of 15.5K census individuals and 600 credit individuals, and Themis measured group and causal discrimination with respect to race and gender, including discrimination-checking experiments with thresholds and pruning comparisons at 99% confidence and 0.05 error bound.
Limitations
- The paper states simplifying assumptions: software is modeled as a black box mapping categorical input characteristics to a categorical output characteristic, and non-categorical inputs/outputs are handled via binning; extending the measures directly to broader data types is left to future work.
- The paper states Themis requires an input schema, and its apparent-discrimination method does not compute confidence because it is only as strong as the developer’s confidence that the test suite or operational profile is representative; apparent discrimination was not directly evaluated.
- Evident from the text: the evaluation is limited to 20 instances of 8 financial-decision systems on two datasets, so the findings may not generalize to other software domains; the paper also leaves integration with combinatorial testing and other automated test-generation improvements to future work.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
0 of 45 cited sources checked · 0 supported · 0 flagged
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“(3) Themis optimizations are effective at producing efficient test suites for measuring discrimination”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesA result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“We find that (1) Themis is effective at discovering software discrimination”
- Also stated in: § ACMReference format: · p. 2 · plain
“Themis is an effective fairness testing tool.”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passes- Also stated in: § ACMReference format: · p. 2 · plain
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“(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”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesA result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“For example, a discrimination-aware decision tree approach trained not to discriminate against gender, B census gender , had a causal discrimination score over 11%: more than 11% of the individuals had the output flipped just by altering the individual's gender.”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesA result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“demonstrating that (i) even when fairness is a design goal, developers can easily introduce discrimination in software”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesA result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“As Section 5.4 will evaluate empirically, pruning leads to, on average, a two-to-three order of magnitude reduction in test suite size.”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesA result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Discrimination is present even in instances designed to avoid discrimination.”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesA result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“For example, for B census gender , the causal score was more than 21 × higher than the group score (11.27% vs. 0.52%).”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesNone of the sources cited for this claim have been checked yet.
In the paper
“Reducing discrimination (CV score) in classifiers [18, 40, 88, 91], as our evaluation has shown, often fails to remove causal discrimination and discrimination against certain groups.”
What the sources say
- Not checked yetcites [18] [18] Toon Calders, Faisal Kamiran, and Mykola Pechenizkiy. Building classifiers with independency constraints. In Proceedings of the 2009 IEEE International Conference on Data Mining (ICDM) Workshops , pages 13-18, Miami, FL, USA, December 2009. - source
- Not checked yetcites [40] [40] Faisal Kamiran, Toon Calders, and Mykola Pechenizkiy. Discrimination aware decision tree learning. In International Conference on Data Mining (ICDM) , pages 869-874, Sydney, Australia, December 2010. - source
- Not checked yetcites [88] source
Load-bearingStrength:boosted· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesA result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Themis was able to discover significant discrimination.”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passes
- It may state two separate claims as one. 32
- Probably not a substantial claim. 30
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“In fact, the discrimination-unaware approaches typically discriminated less than their discrimination-aware counterparts, with the exception of logistic regression.”
Load-bearingStrength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Pruning reduces the number of test cases by, on average, a factor of 2 , 849 for group and 148 for causal discrimination.”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“For example, B census gender discriminates against gender, marital status, and race with a causal score of 77.2%. That means for 77.2% of the individuals, changing only the gender, marital status, or race causes the output of the algorithm to flip.”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“This is the case with B census gender , for example, as its causal discrimination score against gender is 11.2%, but against gender, marital status, and race is 77.2%.”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“show that Themis effectively measures discriminations and produces efficient test suites to do so.”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Themis is effective at finding both group and causal discrimination.”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“For example, Themis needed 3,413 tests to find sets of characteristics that B credit gender discriminated with a score of more than θ = 0 . 7, but only 10 tests when we reduced θ to 0.6.”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
- The authors state this as an assumption: Decreasing the discrimination threshold θ effectively simulates increasing system discrimination.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Given a schema describing valid system inputs, Themis generates discrimination tests automatically and does not require an oracle.”
- Also stated in: § ACMReference format: · p. 2 · plain
“Themis requires a schema for generating inputs, but does not require an oracle.”
- Also stated in: § ACMReference format: · p. 2 · plain
“Unlike prior work that requires manually written tests [79], Themis automatically generates test suites that measure discrimination.”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
- Also stated in: § ACMReference format: · p. 2 · plain
None of the sources cited for this claim could be matched to a paper in Crossref, so there is nothing to check it against.
In the paper
“Amazon.com, Inc. software that determined where to offer free same-day delivery did not explicitly consider race but made racecorrelated decisions because of correlations between race and other input characteristics [36, 52].”
What the sources say
- Not matchedcites [36] [36] David Ingold and Spencer Soper. Amazon doesn't consider the race of its customers. Should it? Bloomberg , April 21, 2016. http://www.bloomberg.com/ graphics/2016-amazon-same-day. -
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
- Not matchedcites [52]
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
- Not matched
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Test caching has low storage overhead and offers significant runtime gains.”
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✗ Based: failsProbably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds. It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The output domain distance function generalizes our work beyond binary outputs.”
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“These approaches do not capture causality and can miss discrimination that our causal approach detects, e.g., when the software discriminates negatively with respect to a group in one settings, but positively in another.”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Themis is effective. Themis is able to (1) verify that many of the software instances do not discriminate against race and gender, and (2) identify the software that does.”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
In the paper
“In forecasting who would reoffend, the software is 'particularly likely to falsely flag black defendants as future criminals, wrongly labeling them this way at almost twice the rate as white defendants; white defendants were mislabeled as low risk more often than black defendants' [5].”
What the sources say
- Not checked yetcites [5] [5] Julia Angwin, Jeff Larson, Surya Mattu, and Lauren Kirchner. Machine bias. ProPublica , May 23, 2016. https://www.propublica.org/article/ machine-bias-risk-assessments-in-criminal-sentencing. -
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“As a result, for example, while the group and apparent discrimination scores penalize software that gives loans to different fractions of individuals of different races, the causal discrimination score penalizes software that gives loans to individuals of one race but not to otherwise identical individuals of another race.”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesProbably not a substantial claim. It may state two separate claims as one.
In the paper
“Our causal fairness definition is designed specifically to be testable, unlike definitions that require probabilistic estimation or knowledge of the future [38].”
What the sources say
- Not checked yetcites [38] [38] Matthew Joseph, Michael Kearns, Jamie Morgenstern, and Aaron Roth. Fairness in learning: Classic and contextual bandits. CoRR , abs/1207.0016, 2012. https: //arxiv.org/abs/1605.07139. - source
Load-bearingStrength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesProbably not a substantial claim. It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The causal discrimination score identifies changing which characteristics directly affects the output.”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesProbably not a substantial claim.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Our definition extends to the more broad categorical input types, reflecting the relative complexity of arbitrary decision software.”
Load-bearingStrength:plain· not yet calibratedTests✓ Asserted: passes✗ Checkable: fails✓ Load-bearing: passes✓ Based: passesProbably not a substantial claim. It may be too vague to check against evidence.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Our definition captures causal relationships between inputs and outputs”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesProbably not a substantial claim.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Themis could have analyzed the software and detected the discrimination prior to deployment.”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✗ Based: failsProbably not a substantial claim. The paper may give no reason why it holds.
In the paper
“Operational profiles [58] describe the input distributions likely to be observed in the field.”
What the sources say
- Not checked yetcites [58] source
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
In the paper
“Our group discrimination score generalizes the CV score to the software domain with more complex inputs.”
What the sources say
- Not checked yetcites [19] [19] Toon Calders and Sicco Verwer. Three naive Bayes approaches for discriminationfree classification. Data Mining and Knowledge Discovery , 21(2):277-292, 2010. - source
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✗ Based: failsProbably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.
None of the sources cited for this claim have been checked yet.
In the paper
“operational profiles can significantly improve the effectiveness of testing by more accurately representing real-world system use [7, 50]”
What the sources say
- Not checked yetcites [7] [7] Andrea Arcuri and Lionel Briand. A practical guide for using statistical tests to assess randomized algorithms in software engineering. In ACM/IEEE International Conference on Software Engineering (ICSE) , pages 1-10, Honolulu, HI, USA, 2011. -
- Not checked yetcites [50] source
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
In the paper
“the use of operational profiles has been shown to more accurately measure system properties, such as reliability [35].”
What the sources say
- Not matchedcites [35] [35] Herman Hartmann. A statistical analysis of operational profile driven testing. In International Conference on Software Quality, Reliability, and Security Companion (QRS-C) , pages 109-116, Viena, Austria, August 2016. -
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
- Not matched
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Themis measures apparent discrimination by either executing a provided test suite, or by generating a test suite following a provided operational profile.”
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Themis can also measure apparent discrimination (Definition 3.8) via a developer-provided test suite or operational profile.”
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✗ Based: failsProbably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“our work does not attempt to remove discrimination but offers developers a tool to identify and measure discrimination”
Load-bearingStrength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesProbably not a substantial claim. It may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“input characteristics may correlate, e.g., education correlates with age, so perturbing some characteristics without perturbing others may create inputs not representative of the real world.”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesProbably not a substantial claim. It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“We have further described Themis, an approach and its open-source implementation for measuring discrimination in software and for generating efficient test suites to perform these measurements.”
Load-bearingStrength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesProbably not a substantial claim. It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Further, focusing on avoiding discriminating against one characteristic may create discrimination against another, e.g., B census gender limits discrimination against gender but discriminates against race with a causal score of 38.40%.”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
In the paper
“FairTest, an implementation of the unwarranted associations framework [79] uses manually written tests to measure four kinds of discrimination scores: the CV score and a related ratio, mutual information, Pearson correlation, and a regression between the output and sensitive inputs.”
What the sources say
- Not checked yetcites [79]
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The B approach was designed to avoid a variant of group discrimination (as are other discrimination-aware approaches), but this design is, at least in some conditions, insufficient to prevent causal discrimination.”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
In the paper
“FairML [1] uses orthogonal projection to co-perturb characteristics, which can mask some discrimination, but find discrimination that is more likely to be observed in real-world scenarios”
What the sources say
- Not checked yetcites [1] [1] Julius Adebayo and Lalana Kagal. Iterative orthogonal feature projection for diagnosing bias in black-box models. CoRR , abs/1611.04967, 2016. -
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Counterfactual fairness is related to causal fairness but can miss some instances of discrimination, e.g., if loan shows preferential treatment for some purple inputs, but at the same time against some other similar purple inputs.”
Strength:plain· not yet calibratedCausal:conditional· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“This experiment demonstrates that Themis effectively discovers discrimination and can test software for unexpected software discrimination effects across a wide variety of input partitions.”
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The work on operational profiles is complementary to ours: Themis uses operational profiles and work on more efficient test generation from operational profiles can directly benefit discrimination testing.”
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“We have formally defined software fairness testing and introduced a causality-based measure of discrimination.”
Load-bearingStrength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesProbably not a substantial claim. It may state two separate claims as one.
None of the sources cited for this claim have been checked yet.
In the paper
“The number of tests needed to evaluate every possible value pair can be significantly smaller than the exhaustive testing alternative since each test can simultaneously contribute to multiple value pairs [22, 44, 80].”
What the sources say
- Not checked yetcites [22] [22] Marcelo d'Amorim, Steven Lauterburg, and Darko Marinov. Delta execution for efficient state-space exploration of object-oriented programs. In International Symposium on Software Testing and Analysis (ISSTA) , pages 50-60, London, UK, 2007. - source
- Not checked yetcites [44] [44] Christian Kästner, Alexander von Rhein, Sebastian Erdweg, Jonas Pusch, Sven Apel, Tillmann Rendel, and Klaus Ostermann. Toward variability-aware testing. In International Workshop on Feature-Oriented Software Development (FOSD) , pages 1-8, Dresden, Germany, 2012. - source
- Not checked yetcites [80] source
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
In the paper
“Counterfactual fairness requires output probability distributions to match for input populations that differ only in the label value of a sensitive input characteristic [51].”
What the sources say
- Not checked yetcites [51] source
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
None of the sources cited for this claim have been checked yet.
In the paper
“The goal of the generated tests is typically finding bugs [29] or generating specifications [21].”
What the sources say
- Not checked yetcites [29] [29] Gordon Fraser and Andrea Arcuri. Whole test suite generation. IEEE Transactions on Software Engineering (TSE) , 39(2):276-291, February 2013. - source
- Not checked yetcites [21] [21] Valentin Dallmeier, Nikolai Knopp, Christoph Mallon, Gordon Fraser, Sebastian Hack, and Andreas Zeller. Automatically generating test cases for specification mining. IEEE Transactions on Software Engineering (TSE) , 38(2):243-257, March 2012. - source
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“our approach generates tests automatically and measures causal discrimination.”
Load-bearingStrength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes? Based: unclearIt may state two separate claims as one.
In the paper
“AdFisher computes a variant of group discrimination, but it could be integrated with Themis and its algorithms to measure causal discrimination.”
What the sources say
- Not checked yetcites [23] [23] Amit Datta, Michael Carl Tschantz, and Anupam Datta. Automated experiments on ad privacy settings. Proceedings on Privacy Enhancing Technologies (PoPETs) , 1:92-112, 2015. - source
Strength:hedged· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Our main goal is to develop a method to process test executions to measure software discrimination, whereas that is not a goal of combinatorial testing.”
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
None of the sources cited for this claim have been checked yet.
In the paper
“Causal relationships in data management systems [54, 55] can help explain query results [57] and debug errors [82-84] by tracking and using data provenance [56].”
What the sources say
- Not checked yetcites [54]
- Not checked yetcites [55]
- Not checked yetcites [57]
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
None of the sources cited for this claim have been checked yet.
In the paper
“Our prior work on testing software that relies on data management systems has focused on data errors [59, 60], whereas this work focuses on testing fairness.”
What the sources say
- Not checked yetcites [59]
- Not checked yetcites [60] source
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“These tools deal with more complex input spaces than Themis, but none of them focus on testing fairness and they require oracles whereas Themis does not need oracles as it measures discrimination by comparing tests' outputs.”
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
- Rests on an assumption: None of the automated test generation tools focus on testing fairness.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“our oracle is more precise and easier to compute, but is only applicable to fairness testing.”
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✗ Based: failsProbably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds. It may state two separate claims as one.
The authors state this as an assumption.
Software can be modeled as a black box that maps input characteristics to an output characteristic.
“First, we define software as a black box that maps input characteristics to an output characteristic.”
The authors state this as an assumption.
Input characteristics and the output characteristic are categorical variables.
“Second, we assume that the input characteristics and the output characteristic are categorical variables, each having a set of possible values (e.g., race, gender, eye color, age ranges, income ranges).”
The authors state this as an assumption.
For the remainder of the paper, software outputs binary decisions.
“For simplicity of exposition, for the remainder of this paper, we assume software outputs binary decisions - a natural domain for fairness testing.”
The paper relies on this without support.
No prior work automatically generates test suites to measure causal discrimination in software.
“To the best of our knowledge, this work is the first to automatically generate test suites to measure causal discrimination in software.”
“We are aware of no prior testing technique, neither automated nor manual, that produces a measure of a software system's causal discrimination.”
The authors state this as an assumption.
Decreasing the discrimination threshold θ effectively simulates increasing system discrimination.
“decreasing θ effectively simulates increasing system discrimination.”
The paper relies on this without support.
No prior work on operational profile testing has measured software discrimination.
“Meanwhile no prior work on operational profile testing has measured software discrimination.”
The paper relies on this without support.
None of the automated test generation tools focus on testing fairness.
“none of them focus on testing fairness”
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“Thus, causal methods are speciically well suited to identifying 1) proxies of sensitive variables as discussed in subsection 2.1; 2) which subgroups are most (un)fairly treated and diferentiate the types of bias exhibited [118]; and 3) provide transparency with respect to how (classiication) decisions were made [143].”
From Fairness in machine learning: A survey (2024) · cites this paper as [113]
In the citing paper
“As a result, a classifier trained without regard to the protected attribute not only fails to remove discrimination, but it can complicate the detection and mitigation of discrimination downstream via in-processing or post-processing techniques [9,10,18,24,25,34,43,51], such as those we describe next.”
From Interventional Fairness: Causal Database Repair for Algorithmic Fairness (2019) · cites this paper as [18]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Reported group score exceeds causal score despite the stated bound
MinorTheorem 4.3 states that the group discrimination score cannot exceed the causal score for the same characteristics. For A census race, the reported gender group score is 8.90%, above the gender causal score of 7.20%. Sampling uncertainty may explain the estimates, but the displayed scores do not satisfy the stated bound.
“| A census race | 2.10% | 1.13% | 8.90% | 7.20% |”
“Theorem 4.3. Let X be an input type and let S be a decision software with input type X . Then for all sets of characteristics X ′ ⊆ X , ˜ d X ′ ( S ) ≤ ⃗ d X ′ ( S ) .”
Citations
Cites
Scikit-learn: Machine Learning in Python
Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Blondel, M., Müller, A., Nothman, J., Louppe, G., et al. (2012) · arXiv · doi:10.48550/arxiv.1201.0490 · Cited by 71,016
Scikit-learn: Machine Learning in Python
(2011) · Journal of Machine Learning Research · doi:10.5555/1953048.2078195 · Cited by 8,539
Dwork, C., Hardt, M., Pitassi, T., Reingold, O., & Zemel, R. (2012) · Proceedings of the 3rd Innovations in Theoretical Computer Science Conference · doi:10.1145/2090236.2090255 · Cited by 3,501 · cited 1×
“nition of discrimination is that a “better” input is never deprived of the “better” output [24].”
BackgroundInternational Conference on Machine Learning (ICML)-2005
津田, 宏. (2006) · 人工知能学会誌 = Journal of Japanese Society for Artificial Intelligence · Cited by 2,558
Causal inference in statistics: An overview
Pearl, J. (2009) · Statistics Surveys · doi:10.1214/09-ss057 · Cited by 2,336 · cited 4×
“Our causal discrimination score is a stronger measure of discrimination, as it seeks out causality in software, measuring the fraction of inputs for which changing specific input characteristics causes the output to change [67].”
BackgroundInfluential(2013) · Cited by 1,126
The oracle problem in software testing: A survey
Barr, E. T., Harman, M., McMinn, P., Shahbaz, M., & Yoo, S. (2015) · IEEE Transactions on Software Engineering · doi:10.1109/tse.2014.2372785 · Cited by 1,104 · cited 1×
“While prior work has tackled the oracle problem [10, 26, 30] typically using inferred pre- and post-conditions or documentation, our oracle is more precise and easier to compute, but is only applicable to fairness testing.”
BackgroundKusner, M. J., Loftus, J. R., Russell, C., & Silva, R. (2017) · arXiv · doi:10.48550/arxiv.1703.06856 · Cited by 995
Arcuri, A., & Briand, L. (2011) · ACM/IEEE International Conference on Software Engineering: Companion Proceedings (ICSE-Companion) · doi:10.1145/1985793.1985795 · Cited by 989 · cited 4×
“Apparent discrimination captures this notion, applying group or causal discrimination score measurement to a subset of the input domain, which can be described by an operational profile [7, 58].”
BackgroundMethodInfluentialFeedback-Directed Random Test Generation
Pacheco, C., Lahiri, S. K., Ernst, M. D., & Ball, T. (2007) · ACM/IEEE International Conference on Software Engineering: Companion Proceedings (ICSE-Companion) · doi:10.1109/icse.2007.37 · Cited by 879 · cited 1×
“Automated testing research has produced tools to generate tests, including random testing, such as Randoop [63, 64], NightHawk [4], JCrasher [20], CarFast [65], and T3 [69]; search-based testing, such as EvoSuite [29], TestFul [9], and eToc [78]; dynamic symbolic execution tools, such as DSC [37], Symbolic PathFinder…”
Method
Flagged first, then by citations
Research works printed but not linked: 8
- Agitar Technologies . AgitarOne. http://www.agitar.com/solutions/products/ automated_junit_generation.html , 2016 . Agitar Technologies. AgitarOne. http://www.agitar.com/solutions/products/ automated_junit_generation.html, 2016.
- Julia Angwin , Jeff Larson , Surya Mattu , and Lauren Kirchner . Machine bias . ProPublica , May 23, 2016 . Julia Angwin, Jeff Larson, Surya Mattu, and Lauren Kirchner. Machine bias. ProPublica, May 23, 2016.
- 10.5555/1792786.1792789 · DOI
- Katelyn Ferral . Wisconsin supreme court allows state to continue using computer program to assist in sentencing . The Capital Times , July 13, 2016 . http://host.madison.com/ct/news/local/govt-and-politics/ wisconsin-supreme-court-allows-state-to-continue-using-computer-program/ article_7eb67874-bf40-59e3-b62a-923d1626fa0f.html. Katelyn Ferral. Wisconsin supreme court allows state to continue using computer program to assist in sentencing. The Capital Times, July 13, 2016. http://host.madison.com/ct/news/local/govt-and-politics/ wisconsin-supreme-court-allows-state-to-continue-using-computer-program/ article_7eb67874-bf40-59e3-b62a-923d1626fa0f.html.
- Rafi Letzter . Amazon just showed us that ‘unbiased’ algorithms can be inadvertently racist . TECH Insider , April 21, 2016 . http://www.techinsider.io/ how-algorithms-can-be-racist- 2016 - 2014 . Rafi Letzter. Amazon just showed us that ‘unbiased’ algorithms can be inadvertently racist. TECH Insider, April 21, 2016. http://www.techinsider.io/ how-algorithms-can-be-racist-2016-4.
- Satya Nadella . The partnership of the future . Slate , June 28, 2016 . http: //www.slate.com/articles/technology/future_tense/2016/06/microsoft_ceo_ satya_nadella_humans_and_a_i_can_work_together_to_solve_society.html. Satya Nadella. The partnership of the future. Slate, June 28, 2016. http: //www.slate.com/articles/technology/future_tense/2016/06/microsoft_ceo_ satya_nadella_humans_and_a_i_can_work_together_to_solve_society.html.
- Aarti Shahani . Now algorithms are deciding whom to hire, based on voice . NPR All Things Considered , March 2015 . Aarti Shahani. Now algorithms are deciding whom to hire, based on voice. NPR All Things Considered, March 2015.
- 10.5555/1792786.1792798 · DOI
Cited by
Sorted by citations
Items 101–200 of 399
Adaptive fairness improvement based on causality analysis
Zhang, M., & Sun, J. (2022) · ACM Joint European Software Engineering Conference and Symposium on the Foundations of Software Engineering (ESEC/FSE) · doi:10.1145/3540250.3549103 · Cited by 38
“These datasets have been used as the evaluation subjects in multiple previous studies [15, 21, 34, 40, 49, 50].”
BackgroundMethodInfluentialTowards Understanding Fairness and its Composition in Ensemble Machine Learning
Gohar, U., Biswas, S., & Rajan, H. (2023) · ACM/IEEE International Conference on Software Engineering: Companion Proceedings (ICSE-Companion) · doi:10.1109/icse48619.2023.00133 · Cited by 47
“Some research in SE has focused on fairness testing to generate test data inputs and uncover fairness violations [1], [14], [58], [59].”
BackgroundXiao, Y., Liu, A., Li, T., & Liu, X. (2023) · ACM SIGSOFT International Symposium on Software Testing and Analysis (ISSTA) · doi:10.1145/3597926.3598099 · Cited by 45
“[30] first formally defined software fairness and discrimination, then proposed a fairness test method named THEMIS, which randomly generates test cases to measure the software discrimination.”
BackgroundMethodInfluentialFairify: Fairness Verification of Neural Networks
Biswas, S., & Rajan, H. (2023) · ACM/IEEE International Conference on Software Engineering: Companion Proceedings (ICSE-Companion) · doi:10.1109/icse48619.2023.00134 · Cited by 47
“While many of the prior works focused on testing ML models [1, 23, 24], more recent works proposed individual fairness testing on NN [8, 25].”
BackgroundResultInfluentialFair Enough: Searching for Sufficient Measures of Fairness
Majumder, S., Chakraborty, J., Bai, G. R., Stolee, K. T., & Menzies, T. (2023) · ACM Transactions on Software Engineering and Methodology · doi:10.1145/3585006 · Cited by 36
“Later in this paper (see §5.1), we make the case that to reduce the space of metrics to be explored, that kind of analytical argument may actually be misleading.”
BackgroundInfluentialSearch-based fairness testing for regression-based machine learning systems
Perera, A., Aleti, A., Tantithamthavorn, C. K., Jiarpakdee, J., Turhan, B., Kühn, L., & Walker, K. (2022) · Empirical Software Engineering · doi:10.1007/s10664-022-10116-7 · Cited by 33
“Themis (Galhotra et al. 2017) is a random test generation technique that targets group and causal discrimination of the SUT.”
BackgroundMethodResultInfluentialMaking fair ML software using trustworthy explanation
Chakraborty, J., Peng, K., & Menzies, T. (2020) · IEEE/ACM International Conference on Automated Software Engineering (ASE) · doi:10.1145/3324884.3418932 · Cited by 33
“ency = 1−1/n.nneiдhbors Õn i=1 |yˆi− Õn j∈N n neiдhbors(xi) yˆj|(1) Prior works in ML software fairness are of mainly two types - a> Testing ML software to find bias- Galhotra et al. created THEMIS[26], a testing-based tool to find software discrimination, focusing on causality in discriminatory behavior. Udeshi et…”
MethodResultEvaluating Fairness Using Permutation Tests
DiCiccio, C. J., Vasudevan, S., Basu, K., Kenthapadi, K., & Agarwal, D. (2020) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/3394486.3403199 · Cited by 34
“There are numerous open-source packages computing fairness metrics including IBM’s AI Fairness 3601, Google’s ML-Fairness-Gym [13], Themis [14], and FairTest [25], though many do not incorporate formal hypothesis testing.”
BackgroundASTRAEA: Grammar-based Fairness Testing
Soremekun, E., Udeshi, S., & Chattopadhyay, S. (2022) · IEEE Transactions on Software Engineering · doi:10.1109/tse.2022.3141758 · Cited by 41
“Fairness in NLP systems requires unique formalization, which distinguishes ASTRAEA from existing works in fairness testing [35], [72].”
BackgroundMethodInfluentialPrivacy Intrusiveness in Financial-Banking Fraud Detection
Găbudeanu, L., Brici, I., Mare, C., Mihai, I.-C., & Șcheau, M. C. (2021) · Risks · doi:10.3390/risks9060104 · Cited by 39
“…(Canillas et al. 2018), sharing such data with other entities (credit/payment institutions) or merchants and direct or indirect consequences on individuals (discrimination (Galhotra et al. 2017) and (Romei and Ruggieri 2014), bias, lack of provision of a service, inaccurate profiles being created).”
BackgroundSocially Responsible AI Algorithms: Issues, Purposes, and ChallengesPreprint
Cheng, L., Varshney, K. R., & Liu, H. (2021) · Journal of Artificial Intelligence Research · doi:10.48550/arxiv.2101.02032 · Cited by 249
“Common methods in fair classification include blinding [32, 93], causal methods [80, 108], transformation [74, 64, 208], sampling and subgroup analysis [37, 65], adversarial learning [75, 211, 175], reweighing [106, 31], and regularization and constraint optimization [111, 17].”
MethodTesting Machine Learning Algorithms for Balanced Data Usage
Sharma, A., & Wehrheim, H. (2019) · IEEE International Conference on Software Testing, Verification and Validation (ICST) · doi:10.1109/icst.2019.00022 · Cited by 32
“[5].”
BackgroundAutomatic Fairness Testing of Neural Classifiers Through Adversarial Sampling
Zhang, P., Wang, J., Sun, J., Wang, X., Dong, G., Wang, X., Dai, T., & Dong, J. S. (2021) · IEEE Transactions on Software Engineering · doi:10.1109/tse.2021.3101478 · Cited by 33
“The Census Income dataset is published in 1996, which is a commonly used dataset in the literature of fairness research [12], [13], [14], [32], [33].”
BackgroundMethodInfluentialVetrò, A., Torchiano, M., & Mecati, M. (2021) · Government Information Quarterly · doi:10.1016/j.giq.2021.101619 · Cited by 42
Towards Formal Fairness in Machine Learning
Ignatiev, A., Cooper, M., Siala, M., Hébrard, E., & Marques-Silva, J. P. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-58475-7_49 · Cited by 36
“Note that this criterion can be seen as a hard version of the causal discrimination measure presented in [27].”
BackgroundThrough the Data Management Lens: Experimental Analysis and Evaluation of Fair Classification
Islam, M. T., Fariha, A., Meliou, A., & Salimi, B. (2022) · ACM SIGMOD International Conference on Management of Data (SIGMOD) · doi:10.1145/3514221.3517841 · Cited by 33
“Individual Discrimination (ID) [32] |Q | |D| , givenQ = {a ∈ D | ∃b : Xa=Xb ∧ Sa≠Sb ∧Ya≠Yb } individual discrimination [0, 1] ID = 0→ completely fair ID = 1→ completely unfair”
BackgroundInfluentialSelf-citationFairness Measures for Machine Learning in Finance
Das, S. R., Donini, M., Gelman, J., Haas, K., Hardt, M., Katzman, J., Kenthapadi, K., Larroy, P., Yilmaz, P., & Zafar, M. B. (2021) · The Journal of Financial Data Science · doi:10.3905/jfds.2021.1.075 · Cited by 31
“Galhotra et al. (2017) call this the “group discrimination score.””
RULER: discriminative and iterative adversarial training for deep neural network fairness
Tao, G., Sun, W., Han, T., Fang, C., & Zhang, X. (2022) · ACM Joint European Software Engineering Conference and Symposium on the Foundations of Software Engineering (ESEC/FSE) · doi:10.1145/3540250.3549169 · Cited by 33
“The aforementioned methods, namely, THEMIS, AEQUITAS, and SG, are mainly designed for traditional machine learning models, such as logistic regression [26], support vector machine [8], and decision tree [9], etc.”
BackgroundMethodInfluentialCompositional Learning and Verification of Neural Network Controllers
Ivanov, R., Jothimurugan, K., Hsu, S., Vaidya, S., Rajeev Alur, A., & Bastani, O. (2021) · ACM Transactions on Embedded Computing Systems · doi:10.1145/3477023 · Cited by 29
“Furthermore, there has also been recent interest in verifying other properties of machine learning systems such as fairness [3, 13, 29], or semantic properties of computer vision via rare event simulation, including applications to safety for self-driving cars [19, 27, 40, 54].”
BackgroundData distribution debugging in machine learning pipelines
Grafberger, S., Groth, P., Stoyanovich, J., & Schelter, S. (2022) · The VLDB Journal · doi:10.1007/s00778-021-00726-w · Cited by 29
Open the Black Box Data-Driven Explanation of Black Box Decision SystemsPreprint
Pedreschi, D., Giannotti, F., Guidotti, R., Monreale, A., Pappalardo, L., Ruggieri, S., & Turini, F. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1806.09936 · Cited by 43
“…which calls for exploring alternative ways to query and audit the black box in the vicinity of the target instance to the purpose of revealing the statistical and causal logic adopted; this can possibly draw from the body of research in active learning and testing, such as fairness testing [18].”
BackgroundDas, S. R., Stanton, R., & Wallace, N. E. (2023) · Annual Review of Financial Economics · doi:10.1146/annurev-financial-110921-125930 · Cited by 28
Integrating social values into software design patterns
Hussain, W., Mougouei, D., & Whittle, J. (2018) · doi:10.1145/3194770.3194777 · Cited by 28
“Galhotra et al. [18] cite many examples of 'biased 'software systems, which act in a way different to the values intended by its designers; supply and demand software pricing systems that unexpectedly led to price gouging on airline tickets for those trying to evacuate from Hurricane Irma – as the…”
BackgroundDatabase Repair Meets Algorithmic Fairness
Salimi, B., Howe, B., & Suciu, D. M. (2020) · ACM SIGMOD Record · doi:10.1145/3422648.3422657 · Cited by 29
“Causal notions of fairness were motivated by the need to address difficulties generated by statistical fairness and assumes an underlying causal model [18, 17, 24, 29, 11].”
BackgroundApplying Human Values Theory to Software Engineering Practice: Lessons and Implications
Ferrario, M. A., & Winter, E. (2022) · IEEE Transactions on Software Engineering · doi:10.1109/tse.2022.3170087 · Cited by 27
“A growing number of researchers, such as Galhotra et al. [31], use automated SE techniques to test the value of ‘fairness’ in systems.”
BackgroundA Combinatorial Approach to Fairness Testing of Machine Learning Models
Patel, A. R., Chandrasekaran, J., Lei, Y., Kacker, R. N., & Kuhn, D. R. (2022) · IEEE International Conference on Software Testing, Verification and Validation Workshops (ICSTW) · doi:10.1109/icstw55395.2022.00030 · Cited by 28
“In our experiments, we use three datasets, namely the Adult Income [31], German-Credit [32], and COMPAS [25] datasets, that are among the most widely used in the fairness testing domain [1][8][10][13].”
InfluentialBetter Automatic Program Repair by Using Bug Reports and Tests Together
Motwani, M., & Brun, Y. (2023) · ACM/IEEE International Conference on Software Engineering: Companion Proceedings (ICSE-Companion) · doi:10.1109/icse48619.2023.00109 · Cited by 37
“, for software fairness [24], [89], [4], [10], and automated oracle generation [69] can synthesize additional tests to improve FL in ways complementary to our approach.”
MethodSelf-citationEnsuring Fairness Beyond the Training DataPreprint
Mandal, D., Deng, S., Jana, S., Wing, J. M., & Hsu, D. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2007.06029 · Cited by 27
“Further afield, our work has similarity with recent work in fairness testing inspired by the literature on program verification [6, 16, 32].”
ResultInfluentialThe Fairness of Credit Scoring Models
Hurlin, C., Perignon, C., & Saurin, S. (2021) · SSRN Electronic Journal · doi:10.2139/ssrn.3785882 · Cited by 26
Verifying Individual Fairness in Machine Learning ModelsPreprint
John, P. G., Vijaykeerthy, D., & Saha, D. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2006.11737 · Cited by 28
“We performed an experiment by running the random testing algorithm (THEMIS) with 50,000 samples and verifi-cation algorithm to compare the time taken by the verifi-cation algorithm and testing for finding counter-example and determine that our algorithm provides NO BIAS before the testing algorithm exhaust the test case…”
MethodJusticia: A Stochastic SAT Approach to Formally Verify Fairness
Ghosh, B., Basu, D., & Meel, K. S. (2021) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v35i9.16925 · Cited by 25
Reducing implicit gender biases in software development: does intergroup contact theory work?
Wang, Y., & Zhang, M. (2020) · ACM Joint European Software Engineering Conference and Symposium on the Foundations of Software Engineering (ESEC/FSE) · doi:10.1145/3368089.3409762 · Cited by 25
“Fairness testing, though not only focusing on gender equality, provides a unified framework to detect and test software systems for potential discriminations, including that against females [30].”
BackgroundProbabilistic Verification of Neural Networks Against Group Fairness
Sun, B., Sun, J., Dai, T., & Zhang, L. (2021) · Lecture notes in computer science · doi:10.1007/978-3-030-90870-6_5 · Cited by 24
FairPrep: Promoting Data to a First-Class Citizen in Studies on Fairness-Enhancing InterventionsPreprint
Schelter, S., He, Y., Khilnani, J., & Stoyanovich, J. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1911.12587 · Cited by 39
“ming environment, but rather incorporate fairness interventions into data-rich machine learning pipelines, while paying close attention to data pre-processing. Another relevant line of work is Themis [10], a software testing framework that automatically designs fairness tests for black-box systems. Generalchallengesin…”
BackgroundAutomated testing of software that uses machine learning APIs
Wan, C., Liu, S., Xie, S., Liu, Y., Hoffmann, H., Maire, M., & Lu, S. (2022) · ACM/IEEE International Conference on Software Engineering: Companion Proceedings (ICSE-Companion) · doi:10.1145/3510003.3510068 · Cited by 26
Remote explainability faces the bouncer problem
Merrer, E. L., & Trédan, G. (2020) · Nature Machine Intelligence · doi:10.1038/s42256-020-0216-z · Cited by 23
“[16], inspired by statistical causality [31], for instance propose to create input datasets for observing discrimination on some specific features by the system under test.”
BackgroundInfluentialAccurate Fairness: Improving Individual Fairness without Trading Accuracy
Li, X., Wu, P., & Su, J. (2023) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v37i12.26674 · Cited by 26
From Transparency to Justification: Toward Ex Ante Accountability for AIPreprint
Malgieri, G., & Pasquale, F. (2022) · SSRN Electronic Journal · doi:10.2139/ssrn.4099657 · Cited by 23
Rating AI systems for bias to promote trustable applications
Srivastava, B., & Rossi, F. (2019) · IBM Journal of Research and Development · doi:10.1147/jrd.2019.2935966 · Cited by 22
Analysing Gender Differences in Building Social Goal Models: A Quasi-Experiment
Gralha, C., Goulão, M., & Araújo, J. (2019) · IEEE International Requirements Engineering Conference (RE) · doi:10.1109/re.2019.00027 · Cited by 20
“Awareness of these gender biases within software systems has increased in recent years [5], [6], and analysing gender differences with software is important.”
BackgroundAutomated Test Generation to Detect Individual Discrimination in AI ModelsPreprint
Agarwal, A., Lohia, P., Nagar, S., Dey, K., & Saha, D. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1809.03260 · Cited by 21
“The existing technique, such as THEMIS (Galhotra, Brun, and Meliou 2017) generates a test suite to determine if and how much individual discrimination is present in the model.”
BackgroundMethodResultInfluentialSoftware Engineering for Fairness: A Case Study with Hyperparameter OptimizationPreprint
Chakraborty, J., Xia, T., Fahid, F. M., & Menzies, T. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1905.05786 · Cited by 21
“Galhotra and his colleagues discussed how to efficiently generate test cases to test for discrimination [10].”
Johnson, B., & Brun, Y. (2022) · ACM/IEEE International Conference on Software Engineering: Companion Proceedings (ICSE-Companion) · doi:10.1145/3510454.3516830 · Cited by 20
“While there is potential for data-driven software to improve our way of life, recent studies suggest that societal biases are in the data we use when training machine learning models, which leads to technological biases [12, 27].”
BackgroundMethodInfluentialSelf-citationGaba, A., Kaufman, Z., Cheung, J., Shvakel, M., Hall, K. W., Brun, Y., & Xiong, C. (2023) · IEEE Transactions on Visualization and Computer Graphics · doi:10.1109/tvcg.2023.3327192 · Cited by 26
“Because fairness is domain-specific [19, 74], and many definitions are mutually incompatible [23], we use one common definition, demographic parity, which requires the distributions of model predictions to be similar for the sensitive groups [19, 25].”
BackgroundMethodSelf-citationCausal testing: understanding defects' root causesPreprint
Johnson, B., Brun, Y., & Meliou, A. (2020) · arXiv (Cornell University) · Cited by 20
Self-citationDocumenting Ethical Considerations in Open Source AI Models
Gao, H., Zahedi, M., Treude, C., Rosenstock, S., & Cheong, M. (2024) · Proceedings of the ... ACM-IEEE International Symposium on Empirical Software Engineering and Measurement/International Symposium on Empirical Software Engineering and Measurement · doi:10.1145/3674805.3686679 · Cited by 28
“Galhotra et al. [30] proposed a testing-based method to measure software discrimination.”
MethodAvoiding the intrinsic unfairness of the trolley problem
Holstein, T., & Dodig-Crnković, G. (2018) · doi:10.1145/3194770.3194772 · Cited by 19
“Including tools that allow automatic discrimination tests, like Themis [22] that generates efficient test suites to measure discrimination, into the software development cycle”
Reduced Products of Abstract Domains for Fairness Certification of Neural Networks
Mazzucato, D., & Urban, C. (2021) · Lecture notes in computer science · doi:10.1007/978-3-030-88806-0_15 · Cited by 19
“It is designed to certify that the classification is independent of the values of the inputs that are considered (directly or indirectly) sensitive to bias [12].”
BackgroundAutomatic Fairness Testing of Machine Learning Models
Sharma, A., & Wehrheim, H. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-64881-7_16 · Cited by 19
“We do not consider THEMIS [8] for our comparison as it has already been shown to be less effective than the other two approaches as stated by Zhang et al. [24].”
InfluentialFairness-aware practices from developers’ perspective: A survey
Voria, G., Sellitto, G., Ferrara, C., Abate, F., Lucia, A. D., Ferrucci, F., Catolino, G., & Palomba, F. (2025) · Information and Software Technology · doi:10.1016/j.infsof.2025.107710 · Cited by 18
Digital Discrimination in Sharing Economy A Requirements Engineering Perspective
Tushev, M., Ebrahimi, F., & Mahmoud, A. (2020) · IEEE International Requirements Engineering Conference (RE) · doi:10.1109/re48521.2020.00031 · Cited by 18
“The goal of this line of research is to propose formal methods for quantifying discrimination and bias in software and develop algorithmic solutions for fairness testing and preservation, especially in Machine Learning systems where the data can be inherently biased [10], [11], [18], [19], [20].”
BackgroundMethodTowards model-based bias mitigation in machine learning
Yohannis, A., & Kolovos, D. S. (2022) · ACM/IEEE International Conference on Model Driven Engineering Languages and Systems (MODELS) · doi:10.1145/3550355.3552401 · Cited by 18
“Themis [22], a bias toolbox, allows automatic generation of tests to measure discrimination in the decisions of a predictive system.”
BackgroundQuoTe : Quality-oriented Testing for Deep Learning Systems
Chen, J., Wang, J., Ma, X., Sun, Y., Sun, J., Zhang, P., & Cheng, P. (2023) · ACM Transactions on Software Engineering and Methodology · doi:10.1145/3582573 · Cited by 21
“Significant efforts have been made in the software engineering community to enhance the quality of DL systems with a focus on threats like adversarial examples (on robustness) [19, 55, 63, 70].”
BackgroundMethodInfluentialAutomated Program Repair, What Is It Good For? Not Absolutely Nothing!
Eladawy, H. M., Goues, C. L., & Brun, Y. (2024) · ACM/IEEE International Conference on Software Engineering: Companion Proceedings (ICSE-Companion) · doi:10.1145/3597503.3639095 · Cited by 18
Self-citationMartino, V. D., & Palomba, F. (2025) · Information and Software Technology · doi:10.1016/j.infsof.2025.107678 · Cited by 26
Faire: Repairing Fairness of Neural Networks via Neuron Condition Synthesis
Li, T., Xie, X., Wang, J., Guo, Q., Liu, A., Ma, L., & Liu, Y. (2023) · ACM Transactions on Software Engineering and Methodology · doi:10.1145/3617168 · Cited by 21
“[20] deines fairness and discrimination and develops THEMIS to generate eicient test suites to measure discrimination.”
BackgroundHigher income, larger loan? monotonicity testing of machine learning models
Sharma, A., & Wehrheim, H. (2020) · ACM SIGSOFT International Symposium on Software Testing and Analysis (ISSTA) · doi:10.1145/3395363.3397352 · Cited by 17
“[14] they aim at a more systematic generation of test inputs.”
BackgroundFix Fairness, Don’t Ruin Accuracy: Performance Aware Fairness Repair using AutoML
Nguyen, G., Biswas, S., & Rajan, H. (2023) · ACM Joint European Software Engineering Conference and Symposium on the Foundations of Software Engineering (ESEC/FSE) · doi:10.1145/3611643.3616257 · Cited by 23
“We conduct an extensive evaluation of Fair-AutoML using 4 widely used datasets in the fairness literature [1, 31, 62] and 16 buggy models collected from a recent study [6].”
BackgroundMethodFairRec: Fairness Testing for Deep Recommender Systems
Guo, H., Li, J., Wang, J., Liu, X., Wang, D., Hu, Z., Zhang, R., & Xue, H. (2023) · ACM SIGSOFT International Symposium on Software Testing and Analysis (ISSTA) · doi:10.1145/3597926.3598058 · Cited by 15
“has been widely adopted in previous work [13, 30].”
BackgroundMethodInfluentialSoftware Fairness Debt: Building a Research Agenda for Addressing Bias in AI Systems
Santos, R. D. S., Fronchetti, F., Freire, S., & Spinola, R. O. (2024) · ACM Transactions on Software Engineering and Methodology · doi:10.1145/3709357 · Cited by 15
“Software fairness refers to the ethical principle and practice of ensuring that software systems, algorithms, and their outcomes are just, equitable, and unbiased across different groups of people, regardless of their characteristics such as race, gender, ethnicity, or socioeconomic status [18].”
Fagbola, T., & Thakur, S. (2019) · 2019 International Conference on Intelligent Informatics and Biomedical Sciences (ICIIBMS) · doi:10.1109/iciibms46890.2019.8991505 · Cited by 9
“…via a backpropagation approach to detect bias in deeplearning based systems (Avanti, Peyton and Anshul, 2017), i. Themis is an open source bias toolbox that automatically create test collections to evaluate discrimination in decisions made by a predictive learning system (Galhotra et al., 2017).”
MethodTwo Kinds of Discrimination in AI-Based Penal Decision-Making
Hübner, D. (2021) · ACM SIGKDD Explorations Newsletter · doi:10.1145/3468507.3468510 · Cited by 19
“The alternative approaches to algorithmic discrimination that they instead pursue mostly refer to conceptions of causal reasoning [6, 19, 24], often employing standards of counterfactual fairness [18, 27].”
BackgroundInfluentialAequeVox: Automated Fairness Testing of Speech Recognition Systems
Rajan, S. S., Udeshi, S., & Chattopadhyay, S. (2021) · Fundamental Approaches to Software Engineering · doi:10.1007/978-3-030-99429-7_14 · Cited by 20
Grammar Based Directed Testing of Machine Learning Systems
Udeshi, S., & Chattopadhyay, S. (2019) · IEEE Transactions on Software Engineering · doi:10.1109/tse.2019.2953066 · Cited by 15
Fairness Through Robustness: Investigating Robustness Disparity in Deep LearningPreprint
Nanda, V., Dooley, S. W., Singla, S., Feizi, S., & Dickerson, J. P. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2006.12621 · Cited by 15
“In such scenarios, there is often concern regarding the fairness of outcomes of the systems [2, 18].”
BackgroundDiverse Data Selection under Fairness ConstraintsPreprint
Moumoulidou, Z., McGregor, A., & Meliou, A. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2010.09141 · Cited by 15
“For example, individual or causal fairness [27] examine differences in treatment of individuals from different groups who are otherwise very similar, but these are not the focus of this work.”
BackgroundSelf-citationBlindspots in Python and Java APIs Result in Vulnerable Code
Brun, Y., Lin, T., Somerville, J. E. T., Myers, E. M., & Ebner, N. C. (2022) · ACM Transactions on Software Engineering and Methodology · doi:10.1145/3571850 · Cited by 18
“Certain aspects of software systems, such as security or fairness [27], [37], [40] are not only difficult for developers to reason about, it can be difficult for end-users to understand and describe the relevant requirements [47].”
Self-citationA social and technical sustainability requirements catalogue
Moreira, A., Araújo, J., Gralha, C., Goulão, M., Brito, I. S., & Albuquerque, D. (2022) · Data & Knowledge Engineering · doi:10.1016/j.datak.2022.102107 · Cited by 14
Understanding Fairness Requirements for ML-based Software
Baresi, L., Criscuolo, C., & Ghezzi, C. (2023) · IEEE International Requirements Engineering Conference (RE) · doi:10.1109/re57278.2023.00046 · Cited by 16
“We do not consider other definitions, like counterfactual [11] or causal ones [12], because they use causal graphs, and do not involve any prediction algorithm.”
BackgroundSearch-based Automatic Repair for Fairness and Accuracy in Decision-making Software
Hort, M., Zhang, J. M., Sarro, F., & Harman, M. (2024) · Empirical Software Engineering · doi:10.1007/s10664-023-10419-3 · Cited by 15
“Tools such as Themis (Galhotra et al. 2017; Angell et al. 2018) and AEQUITAS (Udeshi et al. 2018) are able to generate tests to detect individual discrimination.”
BackgroundNeuFair: Neural Network Fairness Repair with Dropout
Dasu, V. A., Kumar, A., Tizpaz-Niari, S., & Tan, G. (2024) · ACM SIGSOFT International Symposium on Software Testing and Analysis (ISSTA) · doi:10.1145/3650212.3680380 · Cited by 14
Understanding the Reputation Differences between Women and Men on Stack Overflow
Wang, Y. (2018) · Asia-Pacific Software Engineering Conference (APSEC) · doi:10.1109/apsec.2018.00058 · Cited by 15
“Inspired by research in Machine Learning area, fairness test, a newly presented idea of testing software for discrimination, could also help with challenges of inclusiveness although it is designed for detecting multiple types of biases [47].”
MethodMLCHECK– Property-Driven Testing of Machine Learning Classifiers
Sharma, A., Demir, C., Ngomo, A.-C. N., & Wehrheim, H. (2021) · International Conference on Machine Learning and Applications · doi:10.1109/icmla52953.2021.00123 · Cited by 20
Makhija, N., & Gatterbauer, W. (2023) · Proceedings of the ACM on Management of Data · doi:10.1145/3626715 · Cited by 13
“Recent approaches show that explanations benefit a variety of applications, such as ensuring or testing fairness [34, 63, 66] or finding bias [78].”
BackgroundAdversarial Sampling for Fairness Testing in Deep Neural Network
Ige, T., Marfo, W., Tonkinson, J., Adewale, S., & Matti, B. H. (2023) · International Journal of Advanced Computer Science and Applications · doi:10.14569/ijacsa.2023.0140202 · Cited by 13
“Although proposals and conceptual framework had been researched and formulated to address the issue of fairness in ML model [30], [31], [32].”
BackgroundPreFair: Privately Generating Justifiably Fair Synthetic Data
Pujol, D., Gilad, A., & Machanavajjhala, A. (2023) · Proceedings of the VLDB Endowment · doi:10.14778/3583140.3583168 · Cited by 18
MLOps, LLMOps, FMOps, and Beyond
Tantithamthavorn, C. K., Palomba, F., Khomh, F., & Chua, J. J. (2024) · IEEE Software · doi:10.1109/ms.2024.3477014 · Cited by 13
Building Your Own Product Copilot: Challenges, Opportunities, and Needs
Parnin, C., Soares, G., Pandita, R., Gulwani, S., Rich, J., & Henley, A. Z. (2025) · IEEE International Conference on Software Analysis, Evolution and Reengineering (SANER) · doi:10.1109/saner64311.2025.00039 · Cited by 32
Revenäs, Å., Johansson, A., & Ehn, M. (2020) · JMIR Research Protocols · doi:10.2196/20061 · Cited by 15
“RenderX inclusiveness is an important design goal, discrimination is a common deficiency of digital support systems [14].”
BackgroundDo automated program repair techniques repair hard and important bugs?
Motwani, M., Sankaranarayanan, S., Just, R., & Brun, Y. (2018) · ACM/IEEE International Conference on Software Engineering: Companion Proceedings (ICSE-Companion) · doi:10.1145/3180155.3182533 · Cited by 12
Self-citationCoverage-Guided Fairness Testing
Morales, D. P., Kitamura, T., & Takada, S. (2021) · Studies in computational intelligence · doi:10.1007/978-3-030-79474-3_13 · Cited by 13
Data Management for Causal Algorithmic FairnessPreprint
Salimi, B., Howe, B., & Suciu, D. M. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1908.07924 · Cited by 20
“The lack of universality and the impossibility result for fairness definitions based on associational definitions have motivated definitions based on causality [17, 16, 25, 37, 13].”
BackgroundFairness Testing of Machine Learning Models Using Deep Reinforcement Learning
Xie, W., & Wu, P. (2020) · IEEE International Conference on Trust, Security and Privacy in Computing and Communications (TrustCom) · doi:10.1109/trustcom50675.2020.00029 · Cited by 12
“Galhotra et al . proposed THEMIS [9] to evaluate the fairness score of a machine learning model with random test inputs.”
BackgroundReFAIR: Toward a Context-Aware Recommender for Fairness Requirements Engineering
Ferrara, C., Casillo, F., Gravino, C., Lucia, A. D., & Palomba, F. (2024) · ACM/IEEE International Conference on Software Engineering: Companion Proceedings (ICSE-Companion) · doi:10.1145/3597503.3639185 · Cited by 17
“At the same time, automated fairness testing procedures have been defined [25].”
BackgroundAn Empirical Study of Testing Machine Learning in the Wild
Openja, M., Khomh, F., Foundjem, A. T., Jiang, Z. M., Abidi, M., & Hassan, A. E. (2024) · ACM Transactions on Software Engineering and Methodology · doi:10.1145/3680463 · Cited by 16
Capuchin: Causal Database Repair for Algorithmic FairnessPreprint
Salimi, B., Rodriguez, L., Howe, B., & Suciu, D. M. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1902.08283 · Cited by 12
“discrimination, but it can complicate the detection and mitigation of discrimination downstream via in-processing or post-processing techniques [42, 16, 10, 9, 23, 22, 33, 49], which we next describe.”
BackgroundBias and fairness in machine learning and artificial intelligence
Cirillo, D., & Rementeria, M. J. (2022) · Elsevier eBooks · doi:10.1016/b978-0-12-821392-6.00006-6 · Cited by 14
Dehal, R. S., Sharma, M., & Santos, R. D. S. (2024) · IEEE/ACM International Conference on Software Engineering: Software Engineering in Society (ICSE-SEIS) · doi:10.1145/3639475.3640098 · Cited by 11
“These initiatives approach the problem from various angles, including the improvement of software development [3, 20], the implementation of fairness testing to mitigate bias [7], and considerations from a software management standpoint [20, 25].”
BackgroundWhat Is Fair? Defining Fairness in Machine Learning for Health
Gao, J., Chou, B., McCaw, Z. R., Thurston, H., Varghese, P., Hong, C., & Gronsbell, J. L. (2025) · Statistics in Medicine · doi:10.1002/sim.70234 · Cited by 17
Learning to Generate Fair Clusters from Demonstrations
Galhotra, S., Saisubramanian, S., & Shlomo Zilberstein, B. (2021) · Proceedings of the AAAI/ACM Conference on AI Ethics and Society · doi:10.1145/3461702.3462558 · Cited by 11
“• Group fairness ensures the outcome distribution is the same for all groups of interest [14, 20, 43].”
BackgroundSelf-citationDataPrism: Exposing Disconnect between Data and Systems
Galhotra, S., Fariha, A., Lourenço, R., Freire, J., Meliou, A., & Srivastava, D. (2022) · SIGMOD Conference · doi:10.1145/3514221.3517864 · Cited by 18
“Data-driven approaches have been taken for causal-inference-based fault localization [5, 6, 18, 35, 68], software testing [28, 29, 33, 39, 43, 45, 78], and statistical debugging [50, 79].”
MethodSelf-citationTestSGD : Interpretable Testing of Neural Networks against Subtle Group Discrimination
Zhang, M., Sun, J., Wang, J., & Sun, B. (2023) · ACM Transactions on Software Engineering and Methodology · doi:10.1145/3591869 · Cited by 12
“These datasets have been widely used as evaluation subjects in multiple previous studies on fairness [13, 16, 29, 36, 49, 50].”
InfluentialCounterfactual Situation Testing: Uncovering Discrimination under Fairness given the Difference
Ruggieri, S., & Álvarez, J. M. (2023) · ACM Conference on Equity and Access in Algorithms, Mechanisms, and Optimization (EAAMO) · doi:10.1145/3617694.3623222 · Cited by 15
“, [2, 25, 51]) keep the focus on comparing multiple control-test instances for making individual claims, providing evidence similar to that produced by the quantitative tools used in court [38].”
ResultMetrics and methods for a systematic comparison of fairness-aware machine learning algorithmsPreprint
Jones, G. P., Hickey, J. M., Stefano, P. G. D., Dhanjal, C., Stoddart, L. C., & Vasileiou, V. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2010.03986 · Cited by 28
“[22] designed test cases to automatically identify sources of discrimination in black-box decisioning procedures.”
BackgroundMalgieri, G. (2021) · Law and Business · doi:10.2478/law-2021-0003 · Cited by 10
Online Fairness Auditing through Iterative Refinement
Maneriker, P., Burley, C., & Parthasarathy, S. (2023) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/3580305.3599454 · Cited by 21
“Fairness testing [18] provides a notion of causal fairness and generates tests to check the fairness of a given decision-making procedure.”
BackgroundProvably Valid and Diverse Mutations of Real-World Media Data for DNN Testing
Yuan, Y., Pang, Q., & Wang, S. (2024) · IEEE Transactions on Software Engineering · doi:10.1109/tse.2024.3370807 · Cited by 19
“However, they are limited to specific scenarios or data types like traffic scene images [5], [6], [7], [8], [9], [10] or natural language text [11], [12], [13], [14].”
BackgroundMethodInfluentialWang, Z., Zhang, M., Yang, J., Shao, B., & Zhang, M. (2024) · ACM/IEEE International Conference on Software Engineering: Companion Proceedings (ICSE-Companion) · doi:10.1145/3597503.3639181 · Cited by 10
A comprehensive analysis of gender, racial, and prompt-induced biases in large language models
Torres, N., Ulloa, C., Araya, I., Ayala, M., & Jara, S. (2024) · International Journal of Data Science and Analytics · doi:10.1007/s41060-024-00696-6 · Cited by 14
Rango: Adaptive Retrieval-Augmented Proving for Automated Software Verification
Thompson, K., Saavedra, N., Carrott, P., Fisher, K. D., Sanchez-Stern, A., Brun, Y., Ferreira, J. F., Lerner, S., & First, E. (2025) · ACM/IEEE International Conference on Software Engineering: Companion Proceedings (ICSE-Companion) · doi:10.1109/icse55347.2025.00161 · Cited by 14
“Formal languages can be extended to be more expressive, to capture privacy properties [83], data-based properties [59], [60], fairness properties [12], [27], among others.”
Self-citation
Metadata
- Authors
- Galhotra, S., Brun, Y., & Meliou, A.
- Affiliations
- University of Massachusetts Amherst (US)
- Venue
- Proceedings of the 2017 11th Joint Meeting on Foundations of Software Engineering
- Publisher
- ACM
- Published
- 2017-08-21
- Pages
- 498-510
- OpenAlex
- W2730550703
- Cited by
- 405
- References
- 91
- Citations by year
- 2026: 26 · 2025: 55 · 2024: 46 · 2023: 56 · 2022: 50 · 2021: 53 · 2020: 57 · 2019: 37 · 2018: 22 · 2017: 3
- FWCI
- 23.94
- Type
- conference paper
- Language
- en
- Open access
- green
- arxiv.org
- Locations
- ACM Joint European Software Engineering Conference and Symposium on the Foundations of Software Engineering (ESEC/FSE) (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Funders
- National Science Foundation (CCF-1453474, IIS-1453543, CNS-1744471)
- Topics
- Physical Sciences › Computer Science › Software › Software Testing and Debugging Techniques
- Physical Sciences › Computer Science › Artificial Intelligence › Ethics and Social Impacts of AI
- Physical Sciences › Computer Science › Artificial Intelligence › Explainable Artificial Intelligence (XAI)
- Keywords
- algorithmic discrimination, fairness testing, automated test generation, THEMIS, algorithmic fairness, automated decision systems
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
A language model drew the claims from the paper on 2026-10-09, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-09 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
