Characterizing Location-based Mobile Tracking in Mobile Ad Networks
Characterizing Location-based Mobile Tracking in Mobile Ad Networks
复制标题
DOI：
10.1109/cns.2019.8802632
复制
发表时间：
2019-03
期刊：
影响因子：
--
通讯作者：
Boyang Hu;Qicheng Lin;Yao Zheng;Qiben Yan;Matthew Troglia;Qingyang Wang
中科院分区：
文献类型：
--
作者：
Boyang Hu;Qicheng Lin;Yao Zheng;Qiben Yan;Matthew Troglia;Qingyang Wang
如今的移动应用程序通常与第三方广告库打包在一起，以通过用户数据获利。许多移动广告网络利用这些移动应用程序来提取有关用户的敏感实时地理数据，以进行基于位置的定向广告。然而，广告网络大量收集敏感信息引发了严重的隐私问题。不幸的是，基于位置的广告网络的私人数据收集的范围和粒度仍然模糊。在这项工作中，我们提出了一项移动跟踪测量研究，通过使用跨不同地理区域运行的自动化细粒度数据收集工具来表征移动广告网络中基于位置的私人数据收集的严重性和重要性。我们使用在 10 个不同城市运行的 1,100 个流行应用程序对不同的广告网络进行广泛的威胁评估。本研究发现，基于位置的广告数量往往与位置的人口密度呈正相关，不同位置的广告网络的数据收集行为有所不同，并且大多数广告网络都能够收集精确的位置数据。详细分析进一步揭示了地理位置对定向广告跟踪行为的重大影响，以及广告组织在多个应用程序中聚合不同类型的私人用户数据以获得更好的定向广告体验的一个值得注意的安全问题。
Mobile apps nowadays are often packaged with third-party ad libraries to monetize user data. Many mobile ad networks exploit these mobile apps to extract sensitive real-time geographical data about the users for location-based targeted advertising. However, the massive collection of sensitive information by the ad networks has raised serious privacy concerns. Unfortunately, the extent and granularity of private data collection of the location-based ad networks remain obscure. In this work, we present a mobile tracking measurement study to characterize the severity and significance of location-based private data collection in mobile ad networks, by using an automated fine-grained data collection instrument running across different geographical areas. We perform extensive threat assessments for different ad networks using 1,100 popular apps running across 10 different cities. This study discovers that the number of location-based ads tend to be positively correlated with the population density of locations, ad networks' data collection behaviors differ across different locations, and most ad networks are capable of collecting precise location data. Detailed analysis further reveals the significant impact of geolocation on the tracking behavior of targeted ads, and a noteworthy security concern for advertising organizations to aggregate different types of private user data across multiple apps for a better targeted ad experience.