Nuix-Serco Team Wins Navy Contest for Automating Detection of Controlled Unclassified Information

A Nuix and Serco research team has won a U.S. Navy contest aimed at applying artificial intelligence to automate the assessment of controlled unclassified information.

The two-company team was one of three competing groups to win a prize. The work involved training Nuix’s natural language processing system with a data set so it automatically detects CUI in an unstructured text document and assigns it a confidence score.

The Navy seeks to deploy the solution in an internal lab network.

The CUI Automation Challenge was launched by the Naval Surface Warfare Center and is sponsored by the Office of Naval Research.

Oliver Harvey, Nuix’s executive vice president of customer strategy and innovation, described the CUI evaluation process as “laborious” and “open to human error.”

Nuix said in a press release that classification levels enable Department of Defense organizations to simplify information-sharing with partners and allies. The Australia-based software development company added that maintaining trust for information with a classification level is “complex and labor-intensive.”

Results generated by the CUI Automation Challenge could have implications for public and private sector organizations looking to process growing amounts of data, Nuix said.

IBM took first place in the contest followed by the Serco/Nuix team in second place and the Naval Information Warfare Center Pacific Division’s Machine Learning and Natural Language Processing Group in third.