A recent trend on the Web is a demand for higher levels of expressiveness in the mechanisms that mediate interactions such as the allocation of resources, matching of peers, or elicitation of opinions. In this paper, we demonstrate the need for greater expressiveness in privacy mechanisms, which control the conditions under which private information is shared on the Web. We begin by adapting our recent theoretical framework for characterizing expressiveness to this domain. By leveraging prior results, we are able to prove that any increase in allowed expressiveness for privacy mechanisms leads to a strict improvement in their efficiency (i.e., the ability of individuals to share information without violating their privacy constraints). We validate these theoretical results with a week-long human subject experiment, where we tracked the locations of $30$ subjects. Each day we collected their stated ground truth privacy preferences regarding sharing their locations with different groups of people. Our results confirm that i) most subjects had relatively complex privacy preferences, and ii) that privacy mechanisms with higher levels of expressiveness are significantly more efficient in this domain.

25 pages

Return to: SCS Technical Report Collection
School of Computer Science homepage

This page maintained by reports@cs.cmu.edu