arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2611.04510cs.AI
收藏

快速、廉价且公平?评估决策模型

Fast, Cheap, and Fair? Evaluating Decision Models

Amajor SB S.p.A(Amajor SB 股份公司);University of Padova(帕多瓦大学);Human Inspired Technology Research Centre(仿人技术研究中心);University of Trento(特伦托大学);Fondazione Bruno Kessler(布鲁诺·凯斯勒基金会)

Muhammad Uzair Ul Haq, Davide Rigoni, Alessandro Sperduti, Flavia Fabris

AI总结:

本文针对招聘场景的类型化决策模型\nev{}开展系统性公平审计,对比其与噪声基准的偏差影响,评估输出有效性、延迟与成本,并发布相关数据集和代码以实现可复现性。

摘要:

各行业正越来越多地采用AI驱动系统来自动化各类任务,包括人力资源领域的高风险决策,如候选人筛选与招聘。这使得评估其公平性至关重要,尤其是当决策直接影响个人机会时。因此,《欧盟人工智能法案》将招聘AI系统归类为高风险,对其部署施加严格要求。在此背景下,紧凑的“System~1”决策模型提供了一种比通用大语言模型(LLM)更快速、成本更低的替代方案,但其公平性在很大程度上仍未被探索。本文首次对针对招聘应用的类型化决策模型\nev{}开展系统性偏差与公平性审计,研究评估了显性人口统计属性与隐性候选人线索如何影响筛选分数、排名及入围结果,并将这些影响与通过姓名替换和排版变化得到的噪声基准进行比较,同时还评估了输出有效性、延迟及成本,为实现可复现性,研究发布了反事实简历数据集与评估代码。

英文摘要:

Industries are increasingly adopting AI-driven systems to automate a wide range of tasks, including high-stakes decisions in human resources, such as candidate screening and recruitment. This makes evaluating their fairness essential, particularly when decisions directly affect individuals' opportunities. Accordingly, the EU AI Act classifies recruitment AI systems as high-risk, imposing strict requirements on their deployment. In this context, compact ``System~1'' decision models offer a fast and cost-efficient alternative to general-purpose LLMs, yet their fairness remains largely unexplored. This article presents the first systematic bias and fairness audit of \jev{}, a typed-decision model targeting recruitment applications. The study evaluates how explicit demographic attributes and implicit candidate cues influence screening scores, rankings, and shortlisting outcomes. These effects are compared against a noise floor derived from name substitutions and typographical variations. The evaluation also examines output validity, latency, and cost. Counterfactual CV datasets and evaluation code are released for reproducibility.

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。

↑