Craig E. Wills
ABSTRACT
This work seeks to understand what "they" (Web advertisers) actually do with the information available to them. We analyze the ads shown to users during controlled browsing as well as examine the inferred demographics and interests shown in Ad Preference Managers provided by advertisers.
In an initial study of ad networks and a focused study of the Google ad network, we found many expected contextual, behavioral and location-based ads along with combinations of these types of ads. We also observed profile-based ads. Most behavioral ads were shown as categories in the Ad Preference Manager (APM) of the ad network, but we found unexpected cases where the interests were not visible in the APM. We also found unexpected behavior for the Google ad network in that non-contextual ads were shown related to induced sensitive topics regarding sexual orientation, health and financial matters.
In a smaller study of Facebook, we did not find clear evidence that a user's browsing behavior on non-Facebook sites influences the ads shown to the user on Facebook, but we did observe such influence when the Facebook Like button is used to express interest in content. We did observe Facebook ads appearing to target users for sensitive interests with some ads even asserting such sensitive information, which appears to be a violation of Facebook's stated policy.
- Julia Angwin. The web's new gold mine: Your secrets. Wall Street Journal, July 30 2010. http://online.wsj.com/article/SB10001424052748703940904575395073512989404.html.Google Scholar
- Rebecca Balebako, Pedro Leon, Richard Shay, Blase Ur, and Lorrie Faith Cranor. Measuring the effectiveness of privacy tools for limiting behavioral advertising. In Web 2.0 Workshop on Security and Privacy, May 2012.Google Scholar
- Facebook advertising guidelines, Last Revised: March 20, 2012. http://www.facebook.com/ad_guidelines.php .Google Scholar
- Facebook help center: Ad and sponsored stories copy, image, targeting, and destination. http://www.facebook.com/help/?page=245316378826196.Google Scholar
- Fiddler web debugging proxy. http://www.fiddler2.com/fiddler2/.Google Scholar
- Google Ads Preferences. http://www.google.com/ads/preferences/.Google Scholar
- Interest-based advertising: How it works - Google ads preferences, 2010. http://www.google.com/ads/preferences/html/intl/en/about.html.Google Scholar
- Google privacy policy, Last Modified: March 1, 2012. http://www.google.com/intl/en/policies/privacy/.Google Scholar
- Saikat Guha, Bin Cheng, and Paul Francis. Challenges in measuring online advertising systems. In Proceedings of IMC, November 2010. Google ScholarDigital Library
- Balachander Krishnamurthy, Konstantin Naryshkin, and Craig E. Wills. Privacy leakage vs. protection measures: The growing disconnect. In Proceedings of the Web 2.0 Security and Privacy Workshop, pages 1--10, Oakland, CA USA, May 2011.Google Scholar
- Balachander Krishnamurthy and Craig E. Wills. On the leakage of personally identifiable information via online social networks. In WOSN, 2009. Google ScholarDigital Library
- Balachander Krishnamurthy and Craig E. Wills. Privacy diffusion on the web: A longitudinal perspective. In Proceedings of the World Wide Web Conference, 2009. Google ScholarDigital Library
- Jonathan Mayer. Tracking the trackers: Where everybody knows your username, October 2011. http://cyberlaw.stanford.edu/node/6740.Google Scholar
- Franziska Roesner, Tadayoshi Kohno, and David Wetherall. Detecting and defending against third-party tracking on the web. In Symposium on Networked Systems Design and Implementation, April 2012. Google ScholarDigital Library
- Arnold Roosendaal. Facebook tracks and traces everyone: Like this! Technical report, Tilburg Law School, November 2010. Legal Studies Research Paper Series No. 03/2011. http://ssrn.com/abstract=1717563 .Google Scholar
- Craig E. Wills and Can Tatar. Understanding what they do with what they know. Technical Report WPI-CS-TR-12-03, Computer Science Department, Worcester Polytechnic Institute, August 2012. http://www.cs.wpi.edu/~cew/papers/tr12-03.pdf.Google Scholar
Index Terms
- Understanding what they do with what they know
Recommendations
Commentary---When Is Less More, and How Much More? Thoughts on the Psychological and Economic Implications of Online Targeting and Obtrusiveness
In a very intriguing and groundbreaking study, Goldfarb and Tucker [Goldfarb, A., C. Tucker. 2011. Online display advertising: Targeting and obtrusiveness. Marketing Sci.30(3) 389--404] show that online advertising targeting and obtrusiveness boost ...
Exploring ADINT: Using Ad Targeting for Surveillance on a Budget - or - How Alice Can Buy Ads to Track Bob
WPES '17: Proceedings of the 2017 on Workshop on Privacy in the Electronic SocietyThe online advertising ecosystem is built upon the ability of advertising networks to know properties about users (e.g., their interests or physical locations) and deliver targeted ads based on those properties. Much of the privacy debate around online ...
Privacy Regulation and Online Advertising
Advertisers use online customer data to target their marketing appeals. This has heightened consumers' privacy concerns, leading governments to pass laws designed to protect consumer privacy by restricting the use of data and by restricting online ...
Reviews
William Edward Mihalo
This paper describes research associated with systems security. It will be of interest to system administrators and people monitoring network security for an organization. It will also be of interest to people doing research in systems security and people interested in studying online privacy. The authors propose a method for studying browsing behavior and its effect on online privacy. They are particularly interested in the ad preference managers provided by advertisers on four major ad networks: AOL, BlueKai, Google, and Yahoo. The authors regard an ad network as a black box in terms of how it works. They note that an ad network has "inputs that can be controlled and outputs that can be observed." In effect, the authors developed a system to reverse-engineer the perceived logic used by the major ad networks. This paper is the abridged version of a longer paper, which is listed in the references. Unfortunately, the uniform resource locator (URL) listed is a dead link; the correct link for the longer paper is http://www.cs.wpi.edu/~cew/papers/tr1203.pdf. The authors found that "behavioral ads based [on] induced interests were shown as categories in the ad preference manager." They also found some unexpected cases where the interests were not visible. The Google ad network displayed unexpected behavior: "noncontextual ads were shown related to induced sensitive topics regarding sexual orientation, health, and financial matters." The authors did a smaller study with Facebook; if a user uses the "Like" button to express interest, the user's browsing behavior on non-Facebook sites influences the ads shown on Facebook. This study is interesting, but this short version of the results is hard to follow because some explanations and definitions were omitted. The longer version of the paper, which is 24 pages long, is easier to follow. Online Computing Reviews Service
Access critical reviews of Computing literature here
Become a reviewer for Computing Reviews.
Comments