Private and continual release of statistics
Websites such as recommender systems, search engines and social networks commonly publish aggregate statistics about their users to realize valuable social and economic utilities. Moreover, the published statistics are continually updated over time as new data arrive.
Releasing aggregate information about users may seem harmless at first glance. However, previous work has shown that such statistical disclosures can expose sensitive information about an individual user. In this paper, we ask the question -- how can we guarantee the users' privacy when a website must continually publish new statistics as new data arrive? We propose a differentially-private counter mechanism that continually publishes the counts at every time step.
Chan, H.; Shi, E.; Song, D. Private and continual release of statistics. 37th International Colloquium on Automata, Languages and Programming (ICALP); 2010 July 5-10; Bordeaux, France. Berlin: Springer; 2010; LNCS 6199/2010: 405-417.