The Internet has transformed from a small non-profit network into a gigantic infrastructure that creates revenues measured in billions of dollars. Unfortunately, this has created an unprecedented pressure on the end user's privacy by trackers and information aggregators, because a user's profile is used to determine which personalized content (e.g., ads, search results, recommendations, etc.) will be served to the given user. The common goal of the trackers is to gather a more perfect user profile in an attempt to increase revenues from content personalization. Such user profiling can often be misused. For example, signs of price and search discrimination have been recently reported. Others argue that content personalization, which creates the so-called 'filter bubble' effect, has an even more profound impact on the society and in particular on the future of democracy.
The PI proposes endpoint user profile control as a comprehensive approach to the above personalization-induced problems. In the PI's approach, the user has the ability and means, which this project will develop, to explicitly define and implicitly control its profile at all possible trackers at once. The PI will develop and implement endpoint user profile controller as a thin sub-layer that 'sits' between the end-user and endpoint client applications. The controller imprints the desired user profiles by leaving controlled online footprints in the public domain, e.g., by actively visiting web pages from a set of selected topics, sending search queries semantically linked with the user-defined profile etc. This allows users to submit their preferences to all possible trackers they encounter, granting the user a single point of control over their profiles.
Intellectual Merit: The proposed research will address fundamental questions that are key to developing and deploying effective endpoint user profile control. The main challenges are how to effectively control numerous ad trackers at once and how to achieve this in a scalable and effective manner. The PI's focus will be on the control of (i) ad trackers, (ii) web site trackers, (iii) search engine trackers, (iv) social-network trackers, and (v) information aggregators. The PI will develop statistical methods to accurately evaluate the properties of these trackers in order to effectively control them. The proposed techniques range from statistical rule mining to developing comprehensive scoring, noise-reduction, and orthogonal probing algorithms that are necessary for effective control.
Broader Impact: By empowering the user to explicitly and comprehensively control his user profile on the Internet, the system will not only help users regain their online privacy, but also enhance human liberties and promote free and open society. This will be possible to achieve without scrutinizing the online advertising industry, which is effectively funding today's Internet. The proposed system is capable of achieving this goal in a realistic and feasible way, without the adoption of a complex infrastructure. The PI plans to design and disseminate endpoint user profile controller as easy-to-use browser extensions, plug-ins, and mobile applications, which will enable effective endpoint user profile control.
