Chorus: a Programming Framework for Building Scalable Differential Privacy Mechanisms
This addresses the problem of deploying differential privacy in production for organizations like Uber, though it is incremental by building on existing mechanisms.
The paper tackles the scalability gap in differential privacy by introducing Chorus, a framework that integrates with a high-performance DBMS to enable scalable implementations of complex mechanisms, achieving significant scalability improvements on real-world queries at Uber.
Differential privacy is fast becoming the gold standard in enabling statistical analysis of data while protecting the privacy of individuals. However, practical use of differential privacy still lags behind research progress because research prototypes cannot satisfy the scalability requirements of production deployments. To address this challenge, we present Chorus, a framework for building scalable differential privacy mechanisms which is based on cooperation between the mechanism itself and a high-performance production database management system (DBMS). We demonstrate the use of Chorus to build the first highly scalable implementations of complex mechanisms like Weighted PINQ, MWEM, and the matrix mechanism. We report on our experience deploying Chorus at Uber, and evaluate its scalability on real-world queries.