Fairness Auditing in Urban Decisions using LP-based Data Combination

Fairness Auditing in Urban Decisions using LP-based Data Combination
复制标题

使用基于 LP 的数据组合进行城市决策的公平性审计

DOI:
10.1145/3593013.3594118
复制
发表时间:
2023
期刊:
and Transparency
影响因子:
--
通讯作者:
Ohannessian, Mesrob
Ohannessian, Mesrob
中科院分区:
--
文献类型:
--
作者:
Yang, Jingyi;Miller, Joel;Ohannessian, Mesrob

文献摘要

参考文献

相似文献

公平性审核通常需要依靠辅助来源(例如人口普查数据)来了解受保护的属性。为了避免对将这些信息与主要数据源联系起来的总体模型做出假设,最近的一项工作建议找到与这两个来源一致的整个可能的公平性评估范围。尽管很有吸引力,但这种方法的当前形式依赖于严格的分析表达式,并且缺乏处理连续决策的能力,例如城市服务指标。我们表明,在这种情况下,直接调整这些表达式可能会导致松散甚至空洞的结果,特别是在审计决策的公平性方面。如果使用的话,审计将会被认为比应有的更加乐观。我们提出了一种线性规划公式来处理连续决策,通过在通过柯尔莫哥洛夫-斯米尔诺夫距离测量统计奇偶性时找到经验公平范围。该问题的大小与数据点的数量成线性关系并且可以有效解决。我们分析了这种方法,并对由此产生的公平性评估提供有限样本保证。然后,我们将其应用于合成数据和 311 芝加哥城市服务数据,并展示其揭示微小但可检测的公平界限的能力。
Auditing for fairness often requires relying on a secondary source, e.g., Census data, to inform about protected attributes. To avoid making assumptions about an overarching model that ties such information to the primary data source, a recent line of work has suggested finding the entire range of possible fairness valuations consistent with both sources. Though attractive, the current form of this methodology relies on rigid analytical expressions and lacks the ability to handle continuous decisions, e.g., metrics of urban services. We show that, in such settings, directly adapting these expressions can lead to loose and even vacuous results, particularly on just how fair the audited decisions may be. If used, the audit would be perceived more optimistically than it ought to be. We propose a linear programming formulation to handle continuous decisions, by finding the empirical fairness range when statistical parity is measured through the Kolmogorov-Smirnov distance. The size of this problem is linear in the number of data points and efficiently solvable. We analyze this approach and give finite-sample guarantees to the resulting fairness valuation. We then apply it to synthetic data and to 311 Chicago City Services data, and demonstrate its ability to reveal small but detectable bounds on fairness.
DOI: 10.1145/3551624.3555289
发表时间: 2022
期刊: and Optimization
影响因子: --
作者:
Singh, Shubham;Shah, Bhuvni;Kanich, Chris;Kash, Ian A.
通讯作者: Kash, Ian A.
DOI: --
发表时间: 2020
期刊: Neural Information Processing Systems
影响因子: --
作者:
S. Wang;Wenshuo Guo;H. Narasimhan;Andrew Cotter;Maya R. Gupta;Michael I. Jordan
通讯作者: Michael I. Jordan
DOI: --
发表时间: 2008
期刊:
影响因子: --
作者:
Martha C. Andrews;K. Kacmar;Gerald L. Blakely;Neil S. Bucklew;Carolina Wilmington
通讯作者: Carolina Wilmington