Privacy-preserving algorithms for data storage, retrieval, and dissemination

Giulia Fanti1, Nihar Shah, Rashmi Vinayak and Kannan Ramchandran

Freedom of speech depends critically on being able to consume and share information without fear of repercussion. Modern communication tools may threaten such freedom of speech by revealing user data and metadata to powerful adversaries. We consider various tools for storing, accessing, and disseminating information with provable privacy guarantees. For example, private information retrieval and search tools can be used to search public databases without revealing one's query to the server. We consider algorithmic design and fundamental limits on the communication complexity of such algorithms. Anonymous messaging, on the other hand, can be used to post a message to a social network without revealing one's identity to state-level adversaries. We propose algorithms that provably prevent such adversaries from statistically inferring the source of a message.

Figure 1
Figure 1: Privacy-preserving algorithms

1UIUC

More information: http://www.eecs.berkeley.edu/~kannanr/project_security.html