DOT harnesses machine learning for regulations

The Department of Transportation has built a machine-learning and data analytics-backed dashboard that helps its legal staff manage its vast field of regulatory data, the agency's chief data officer said.

In one of its latest projects leveraging big data, DOT turned machine learning and big data analytics on regulatory data, CDO Daniel Morgan said in remarks at an AFCEA Bethesda breakfast on Dec. 17.

"You all know the Federal Register takes a snapshot of federal regulation on an annual basis, which means you can parse it and create statistics for your lawyers so they can understand complex regulations -- such as which have the most conditions or the most exceptions," he said.

The capability, he said, was requested by the agency's lawyers who deal with the huge volume of regulations, including those from component agencies charged with regulating vehicles, drivers, public transportation systems, highways, railroads, pipelines and hazardous materials transport.

To create the dashboard, said Morgan, the agency combined cloud services, a data visualization tool and "a little elbow grease."

The dashboard also incorporated open source software from the Mercatus Center, George Mason University's market-oriented policy and regulatory think tank, Morgan said in remarks after his presentation. The dashboard, he said, could be adapted to a few other agencies that have heavy regulatory compliance burdens.