Privacy Integrated Data Stream Queries

Research on differential privacy is generally concerned with examining data sets that are static. Because the data sets do not change, every computation on them produces “one-shot” query results; the results do not change aside from random-ness introduced for privacy. There are many circumstances, however, where this model does not apply, or is simply in-feasible. Data streams are examples of non-static data sets where results may change as more data is streamed. Theo-retical support for differential privacy with data streams has been researched in the form of differentially private stream-ing algorithms. In this paper, we present a practical frame-work for which a non-expert can perform differentially pri-vate operations on data streams. The system is built as an extension to PINQ (Privacy Integrated Queries), a differen-tially private programming framework for static data sets. The streaming extension provides a programmatic interface for the different types of streaming differential privacy from the literature so that the privacy trade-offs of each type of algorithm can be understood by a non-expert programmer