arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2205.08809cs.SE
收藏

软件公平性:分析与综述

Software Fairness: An Analysis and Survey

SnT, University of Luxembourg(卢森堡大学科学与技术研究中心)

Ezekiel Soremekun, Mike Papadakis, Maxime Cordy, Yves Le Traon

AI总结:

本文综述了软件公平性分析的现状,分析了164篇相关文献,揭示了公平性衡量、任务类型、分析方法及数据集等方面的不足,指出需进一步研究交叉/序列偏差、政策偏差处理及人机协作的公平性缓解。

摘要:

在过去十年中,研究者将公平性作为软件属性进行研究。特别是如何工程化公平软件系统?这包括指定、设计和验证公平性属性。然而,针对偏差作为软件工程问题的现有工作景观不清晰,即分析基于学习的软件公平性的技术和研究。本文提供了软件公平性分析的现状综述。为此,我们收集、分类并深入分析了164篇研究基于学习的软件系统的公平性文献。具体而言,我们研究了评估的公平性度量、研究的任务、公平性分析的类型、所提出方法的主要思想以及访问级别(例如黑盒、白盒或灰盒)。我们的发现包括以下几点:(1)公平性问题(如公平性规范和需求工程)研究不足;(2)公平性度量如条件、顺序和交叉公平性研究不足;(3)无结构数据集(如音频、图像和文本)几乎未用于公平性分析;(4)软件公平性分析技术几乎不采用白盒、在处理机器学习(ML)分析方法。总之,我们观察到若干开放挑战,包括需要研究交叉/序列偏差、基于政策的偏差处理以及人机协作的社会技术偏差缓解。

英文摘要:

In the last decade, researchers have studied fairness as a software property. In particular, how to engineer fair software systems? This includes specifying, designing, and validating fairness properties. However, the landscape of works addressing bias as a software engineering concern is unclear, i.e., techniques and studies that analyze the fairness properties of learning-based software. In this work, we provide a clear view of the state-of-the-art in software fairness analysis. To this end, we collect, categorize and conduct an in-depth analysis of 164 publications investigating the fairness of learning-based software systems. Specifically, we study the evaluated fairness measure, the studied tasks, the type of fairness analysis, the main idea of the proposed approaches, and the access level (e.g., black, white, or grey box). Our findings include the following: (1) Fairness concerns (such as fairness specification and requirements engineering) are under-studied; (2) Fairness measures such as conditional, sequential, and intersectional fairness are under-explored; (3) Unstructured datasets (e.g., audio, image, and text) are barely studied for fairness analysis; and (4) Software fairness analysis techniques hardly employ white-box, in-processing machine learning (ML) analysis methods. In summary, we observed several open challenges including the need to study intersectional/sequential bias, policy-based bias handling, and human-in-the-loop, socio-technical bias mitigation.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑