[论文解读] Discrimination through optimization: How Facebook's ad delivery can lead to skewed outcomes

Muhammad Ali, Piotr Sapieżyński+4 more|arXiv (Cornell University)|Apr 3, 2019

Consumer Market Behavior and Pricing参考文献 32被引用 74

一句话总结

本文表明，Facebook 的广告投放可能会基于广告预算和创意内容扭曲谁看到就业和住房广告，即使目标设定是包容性的，也暴露了平台优化和市场效应如何推动歧视性结果的机制。

ABSTRACT

The enormous financial success of online advertising platforms is partially due to the precise targeting features they offer. Although researchers and journalists have found many ways that advertisers can target---or exclude---particular groups of users seeing their ads, comparatively little attention has been paid to the implications of the platform's ad delivery process, comprised of the platform's choices about which users see which ads. It has been hypothesized that this process can "skew" ad delivery in ways that the advertisers do not intend, making some users less likely than others to see particular ads based on their demographic characteristics. In this paper, we demonstrate that such skewed delivery occurs on Facebook, due to market and financial optimization effects as well as the platform's own predictions about the "relevance" of ads to different groups of users. We find that both the advertiser's budget and the content of the ad each significantly contribute to the skew of Facebook's ad delivery. Critically, we observe significant skew in delivery along gender and racial lines for "real" ads for employment and housing opportunities despite neutral targeting parameters. Our results demonstrate previously unknown mechanisms that can lead to potentially discriminatory ad delivery, even when advertisers set their targeting parameters to be highly inclusive. This underscores the need for policymakers and platforms to carefully consider the role of the ad delivery optimization run by ad platforms themselves---and not just the targeting choices of advertisers---in preventing discrimination in digital advertising.

研究动机与目标

• 评估仅凭 Facebook 的广告投放是否就能在不考虑目标参数的情况下，对不同人口群体产生偏斜投放。
• 量化预算、广告创意和图片对性别与种族维度投放偏斜的影响。
• 调查自动相关性分类是否会引发初始投放偏斜。
• 展示就业和住房广告的现实世界影响及监管方面的考量。

提出的方法

• 以受控定位、变动预算和创意，执行数十个 Facebook 广告活动。
• 使用 Facebook Marketing API 按人口维度以两分钟间隔收集投放统计数据。
• 开发方法，利用 DMA 级位置代理和公开可用的选民种族数据来推断种族投放。
• 在保持定位不变的情况下，比较不同预算水平、广告创意内容和图片要素下的投放偏斜。
• 对真实的就业和住房广告进行实验，以观察受保护群体代表性的偏斜。

实验结果

研究问题

• RQ1 即使目标设定中性或包容，市场/优化效应是否也会导致广告投放偏斜？
• RQ2 预算、创意内容和图像如何影响投放受众的人口统计构成？
• RQ3 自动相关性评估是否会在不依赖用户互动的情况下对初始投放偏斜产生影响？
• RQ4 在相同定位下，现实世界的就业和住房广告是否显著存在投放人口统计偏斜？

主要发现

• 仅因市场效应就会发生投放偏斜，受众构成随预算变化而变化（例如，极低预算时男性占比>55%，高预算时男性占比<45%）.
• 广告创意内容显著影响投放偏斜，某些刻板印象在相同定位下驱动>80%男性投放或>90%女性投放。
• 广告图片单独影响投放；更换标题/文本/图片会改变投放模式，表明图像驱动的相关性信号。
• Facebook 的自动图片分类可能自广告运行之初就促成偏斜，因为视觉上无差异的图片也会产生不同的投放。
• 现实中的就业和住房广告在相同定位下也显示出显著偏斜，例如某些职位广告白人占比72%、男性占比90%；或收银员广告观众85%为女性；住房广告在某些变体中黑人占比>72%，在其他变体中>51%。
• 研究发现投放优化加市场效应可能在超出定位控制的范围内造成歧视性结果，促使监管和平台层面的考量。