Content by Joseph Near, David Darais
Differential Privacy Bugs and Why They’re Hard to FindDifferentially private programs provide randomized outputs, and privacy bugs aren’t detectable by observing output values
Tue, 06/15/2021 - 12:02
In previous articles we have explored what differential privacy is, how it works, and how to answer questions about data in ways that protect privacy. All of the algorithms we’ve discussed have been demonstrated via mathematical proof to be…
Summation and Average Queries: Detecting Trends in Your DataThe importance of upper limits and clipping
Tue, 05/11/2021 - 12:03
In our last article, we discussed how to determine how many people drink pumpkin spice lattes in a given time period without learning their identifying information. But say, for example, you would like to know the total amount spent on pumpkin spice…
Counting Queries: Extracting Key Business Metrics From DatasetsHow to achieve differential privacy with common counting queries
Wed, 04/21/2021 - 12:03
How many people drink pumpkin spice lattes in October, and how would you calculate this without learning specifically who is drinking them, and who is not? Although they seem simple or trivial, counting queries are used extremely often. Counting…
Threat Models for Differential PrivacyA look at central, local, and hybrid models
Mon, 04/12/2021 - 12:03
It’s not so simple to deploy a practical system that satisfies differential privacy. Our example in the last post was a simple Python program that adds Laplace noise to a function computed over the sensitive data. For this to work in practice, we’d…
Differential Privacy for Privacy-Preserving Data AnalysisNew blog series from NIST seeks to fill gaps in its Privacy Framework
Thu, 03/25/2021 - 12:03
Does your organization want to aggregate and analyze data to learn trends, but in a way that protects privacy? Or perhaps you are already using differential privacy tools, but want to expand (or share) your knowledge? In either case, NIST’s blog…