|Google Open Sources Differential Privacy Library|
|Written by Kay Ewbank|
|Friday, 13 September 2019|
Google has released an open source version of its differential privacy library, which is used for some of Google’s core products. To make the library easy for developers to use, Google has concentrated on features that it says can be particularly difficult to execute from scratch, like automatically calculating bounds on user contributions.
Differential privacy provides a way to make information available from a dataset without making it obvious whether data from a specific individual is in the dataset. For example, analysts could show changing health trends while keeping the underlying medical records private, or the aggregated results of a survey could be used without the individual responses being revealed, even to the researchers doing the analysis.
Google has been working on differentially-private techniques since the release of RAPPOR for Chrome in 2014. RAPPOR (Randomized Aggregatable Privacy-Preserving Ordinal Response), is used to work on Chrome's security, bugs, and overall user experience from user responses without including anything that would identify a particular user. Google has also used differentially private methods to provide data such as how busy a business is over the course of a day or how popular a particular restaurant’s dish is in Google Maps.
The new library offers statistical functions so developers can compute counts, sums, averages, medians, and percentiles using the library. It also includes an extensible ‘Stochastic Differential Privacy Model Checker library’ that can be used to check you're not exposing information that would identify specific users.
The developers say they've designed the library so that it can be extended to include other features such as additional mechanisms, aggregation functions, or privacy budget management.
or email your comment to: firstname.lastname@example.org
|Last Updated ( Friday, 13 September 2019 )|