Summarizing — Adnostic: Privacy Preserving Targeted Advertising

Photo by Austin Distel on Unsplash

Check my original story on medium

Note: This was my first attempt of paper summarizing. Please excuse my mistakes if any .

This article summarizes the paper Adnostic: Privacy Preserving Targeted Advertising. Here is the link to the - original paper.

Summary

Online behavioral advertising (OBA) is a practice of tracking customers online over time in order to get an idea about their preferences and interests. This customer information is then used for selecting of tailoring ads to present them. But, this is a security issue as it infringes user privacy. The paper proposes a practical solution, which enables user targeting without directly affecting their privacy. Their system called Adnostic, is implemented as a browser extension. It runs an algorithm (behavioral targeting algorithm) on browser, by processing browser’s history database to determine user interests. As user browses more and more, the interest database is continuously updated. The results are stored locally in the browser and are used to choose the ads to display for that user. Results are not communicated outside, as long as user doesn’t click on the ad. The paper discusses complications in their system(Adnostic). One of which is billing or accounting.

Ad networks must bill the correct advertiser without knowing which ad was displayed to the user”[1].

To solve this problem, paper proposes the cryptographic billing system.

Contribution

The paper proposes Adnostic, which enables targeting users without affecting their privacy. The motivation behind Adnostic is to compliment existing behavioral advertising infrastructure and not replace it. To solve the accounting or billing problem, the paper proposes cryptographic billing system.

Strengths

I think that the solution is good if it is accepted widely. I like that they considered possible objections with importance and tried to address them. Also, Adnostic is better than the existing solutions (that I know of, and what this paper mentions)and serves its purpose.

Limitations

I think the primary concern is that, as it affects ad rendering time and page loading time, the delay is a lot. Specifically on the websites having a lot of banners. To me, their solution seems unrealistic, as it is impractical to download all of the ads from heavy websites (and multiple). Also, the solution depends on the involvement of ad-networks, it must be widely accepted to be effective.

Future Work

Future work can be done to prevent fraudulent click (as their system will fail with active content, such as flash) and social engineering attacks. If such system is to be implemented, then it must be wide spread (to aware or convince), as it must be widely accepted (to be effective). Also the performance problem must be addressed as soon as possible.

What do you think about the strengths and weaknesses? Did I miss anything? Also, did you like their approach? Please let me know what you think in the comments. 😃

I am an individual writer just starting out. Also, I am not involved in any partner programs to get paid. Buy me a ☕️ on_ paypal if you like my work! 

https://paypal.me/ruturajkvaidya?locale.x=en_US

[1] Link to the original paper: https://crypto.stanford.edu/adnostic/
Adnostic: Privacy Preserving Targeted Advertising
Vincent Toubiana, Arvind Narayanan, Dan Boneh, Helen Nissenbaum, Solon Barocas 
Appeared at NDSS 2010

You can install the firefox extension here.