Quantity vs. Quality: Evaluating User Interest Profiles Using Ad Preference Managers

Abstract

Widely reported privacy issues concerning major online advertising platforms (e.g., Facebook) have heightened concerns among users about the data that is collected about them. However, while we have a comprehensive understanding who collects data on users, as well as how tracking is implemented, there is still a significant gap in our understanding: what information do advertisers actually infer about users, and is this information accurate? In this study, we leverage Ad Preference Managers (APMs) as a lens through which to address this gap. APMs are transparency tools offered by some advertising platforms that allow users to see the interest profiles that are constructed about them. We recruited 220 participants to install an IRB approved browser extension that collected their interest profiles from four APMs (Google, Facebook, Oracle BlueKai, and Neilsen eXelate), as well as behavioral and survey data. We use this data to analyze the size and correctness of interest profiles, compare their composition across the four platforms, and investigate the origins of the data underlying these profiles.

Oh, the Places You've Been! User Reactions to Longitudinal Transparency About Third-Party Web Tracking and Inferencing

“…This outcome also aligns with real-world tracking. Prior work has documented the poor accuracy of behavioral profiles built by online advertisers [74,89], with one study finding only 27% of inferences were strongly relevant [9]. At least 40% of attributes sold by data brokers may be inaccurate [93].…”

“…For example, Google's Ad Settings (Figure 1b) lists some estimated interests and gives vague explanations of how they were chosen. These dashboards have been shown to be incomplete [89,99], misleading [4], and potentially inaccurate [9,21]. They have also been used to show discrimination in advertising [20] and targeting on sensitive topics [48,99].…”

Awakening the Web's Sleeper Agents: Misusing Service Workers for Privacy Leakage

“…and 507 respectively). While the non-sensitive categories may not directly reveal private information about the user, they are typically part of Ad Preference Manager profiles [17] and can be leveraged by advertisers for user targeting [30]. Moreover, websites that are susceptible to fine-grained history sniffing (see Section VI-C) could enable the inference of sensitive data.…”

Auditing Offline Data Brokers via Facebook's Advertising Platform

“…Finally, a number of recent studies have examined the accuracy of information revealed by the transparency mechanisms of online advertising platforms. While one study [50] found that participants' age and gender as revealed by Google's transparency mechanism were accurate for 65% -74% of participants, and had missing values for 18%-29% of participants; another study [14] found that over 52% of participants reported less than half of their information as listed by the transparency mechanisms of Google, Facebook, and a small broker Nielsen eXelate as "relevant". Finally, 27% of users in a recent report [1] found their information revealed by Facebook's transparency mechanism inaccurate.…”