Big data's big hurdle: federal policy

Big data has all the promise in the world to help agencies glean otherwise impossible insights from massive repositories of data, but some believe federal policies thus far have blunted big data’s potential in the public sector.

"Policy questions are always a potential derailer," said Josh Sullivan, a data scientist and vice president at Booz Allen Hamilton.

"We’ve worked with a number of big agencies on practical pilots and nationally enterprised adoptions," Sullivan told FCW. "Inevitably, security policy, data protection policies and sharing policies were all antibodies to all the technology."

Sullivan said that some intelligence agencies have exceled out of necessity at finding innovative solutions in big data that navigate policy, developing mature big data models that combine IT efficiencies, "smart" data and cloud analytics to create a portfolio on which data-driven decisions are made.

But most other federal agencies, Sullivan said, are content to reach the point at which they’ve jumped into the cloud and virtualized some of their IT infrastructure.

They are reacting to the big data buzz with outdated, old-hat polices rather than proactively embracing potentially huge financial gains and insights through policy that outlines questions that big data can actually answer.

"Until people start saying big analytics instead of big data, I don’t know if we’ll see the government realizing all the latent patterns there for us to discover," Sullivan said. Agency executives can start framing outlines by "writing down 10 game-changing questions you cannot answer today that would substantially move the needle on your mission," he added.

Technology changes fast, said Mark Forman, former administrator for e-government and IT at the Office of Management and Budget, but said broader federal polices aren’t changing with them.

"Laws relating to privacy haven’t kept up with the changes in data as a result of social media and the Internet," said Forman, who believes decision makers will one day base major decisions on insights from structured and unstructured big data.

Forman said the Privacy Act, enacted in 1974 to keep the government from using personal citizen data for reasons not in the public’s best interest, could stand a 21st century refresh.

Some, he said, thought the Bush-era Patriot Act might be the catalyst for such a change, but the controversy over the counterterrorism measure eventually quieted.

Yet big data, and particularly the ability to collect huge bits of unstructured data, has changed how the government gathers information. What used to be a linear approach to information gathering is now a "topsy-turvy" approach where mined data can infringe on existing privacy laws.

"It’s hard to scrape off privacy-related information," Forman said. "There is a lot of research going on right now in understanding how we modify privacy laws and paperwork burden laws in a totally different era, and not just in the United States. The question is, what is the catalyst for this?"

Aubrey Vaughan, managing director of the public sector at Oversight Systems, a big data analytics software provider, said unfortunately no "silver bullet exists" to solve the policy dilemma short of a group of feds championing the effort.

"I think we are in the infancy stages of big data, a lot of agencies are being told a lot of different things about big data, and it is a buzzword," Vaughan said. "Some agencies are putting the emphasis now on analytics, and efficiencies are driving it more to the surface and creating more groundswell in the government to do more now."