PipelineDP4j, an ‘out-of-the-box’ solution for analyzing data sets in Apache Beam and Apache Spark in a privacy-preserving way, is intended to be usable by all developers.
Google has released PipelineDP4j, a differential privacy solution that allows developers to execute highly parallelizable computations using Java as the baseline language, while providing guarantees that personal information is kept private and secure.
PipelineDP4j opens the door for new applications of differential privacy by reducing the barrier of entry for developers already working in Java, Google said. The software executes on the JVM, can be used with Java, Kotlin, or Scala, and supports distributed data processing frameworks such as Apache Beam and, coming soon, Apache Spark.
PipelineDP4j is intended for use by all developers, regardless of differential privacy expertise, Google said. Differential privacy is a PET (privacy-enhancing technology) serving as a framework for analysis of data sets in a privacy-preserving way to ensure that personal information is never released, according to Google.
Announced October 31, PipelineDP4j is an evolution of work done with OpenMined, which builds open source privacy software. Between the open-source differential privacy library and this JVM release, Google said it now covers some of the most popular languages — Python, Java, Go, and C++ — and potentially more than half of all developers worldwide.
PipelineDP4j relies on the lower-level building blocks from the differential privacy library and combines them into an out-of-the-box solution that takes care of the steps essential to differential privacy, including noise addition, partition selection, and contribution bounding, Google said.
