Big Data, privacy, law enforcement | Homeland Security Newswire

Big Data & privacyKeeping Big Data safe

Published 7 May 2018

NIST has announced the Unlinkable Data Challenge, created to help the public safety community conduct research using data gathered with personal digital devices and taken from large databases such as driver’s license and health care records. Much of this data includes personal information that can be used to identify its source. Exposing this data risks those individuals’ privacy, but the inability to share it impedes research in many fields, including thwarting crime, fighting fires and slowing the spread of epidemics.

The National Institute of Standards and Technology (NIST) is requesting the public’s help in making the personal information in large databases safe for scientists to use—without risking the privacy of the individuals behind it.

NIST has announced the Unlinkable Data Challenge, created to help the public safety community conduct research using data gathered with personal digital devices and taken from large databases such as driver’s license and health care records. Much of this data includes personal information that can be used to identify its source. Exposing this data risks those individuals’ privacy, but the inability to share it impedes research in many fields, including thwarting crime, fighting fires and slowing the spread of epidemics.

The key to unleashing the data’s power for the public safety community lies in finding automated ways to effectively “de-identify” personal information while maintaining the data’s analytic value. The goal of the challenge is to create these methods, which will help the public safety community make better decisions while protecting the public from data leaks and cyberattacks.

The challenge will have three phases, and $190,000 of total prize money will be split among them. The first phase asks competitors to propose an overall conceptual approach to de-identifying a data set. The subsequent two phases will involve developing and refining the algorithms to implement the approach.

Submissions for the first phase close on 26 July 2018. Complete details are here.