Tracking the tech that will make government better

Crowdsourcing, fraud detection, and open data tools were touted at a recent Senate hearing.

Gov 2.0 Summit, 2010Will crowdsourcing and next-generation data mining tools enable the federal government to find innovative solutions to grand challenges and reduce fraud?

Last week, the Senate Committee on Homeland Security and Governmental Affairs Subcommittee on Federal Financial Management, Government Information, Federal Services, and International Security held a hearing on “Transforming
Government Through Innovative Tools and Technology
” that looked at
the potential for technology to improve government transparency and
accountability. The first part of the hearing featured testimony from
Daniel Werfel, controller of the Office of Federal Financial
Management within the Office of Management and Budget, and Earl
Devaney, chairman of the Recovery Accountability and Transparency
Board (RATB). You can view their written testimony and archived webcast at Senate.gov.

Riley Crane, a post-doc at the Media Laboratory Human Dynamics
Group at the Massachusetts Institute of Technology, shared one of the
most successful examples of crowdsourcing in history: the strategy
that led to the MIT balloon
team’s
victory
in the DARPA Network
Challenge
.

“For the first time, we can bridge the gap between online and the
real world,” testified Crane. A challenge “thought impossible by the
intelligence community using traditional techniques” was solved in 8
hours and 52 minutes, said Crane. “We leveraged the problem-solving
capabilities of the participants,” said Crane, and “built the
infrastructure that allowed others to solve the problem for us.” As
Brian Ahier pointed out on his blog
on healthcare IT
, Crane praised Tim O’Reilly’s “government as a
platform
” concept and Gov 2.0 principles in his Senate
testimony:

More on the potential of crowdsourcing and how open data analysis is improving fraud detection after the jump.

Tools for government transparency

Devaney asserted that stimulus
tracking technology is the model for government transparency
. As
Gautham Nagesh reported for The Hill, Devaney said that “the amount of fraud he has seen in the Recovery Act is significantly below what he would expect on a program of its size. He argued that the transparency inherent in the program has acted as a deterrent to
scam artists, keeping fraud down to a minimum. He said the RATB uses software to identify risk factors associated with particular awards, contractors or grant recipients, then refers the information to specific inspector generals for further investigation.”

And as Aliya Sternstein reported for Nextgov, “compu-forensics
saved stimulus funds
and Devaney from some pointed questions by
John McCain on how the RAT Board‘s data analytics has saved money or resulted in convictions. Devaney
said that there are more than 350 ongoing criminal investigations into
fraud and abuse of Recovery Act funds, and, while none have resulted
in any convictions yet, he expects to see resolution on some of
them within six months.

Devaney also noted that the Recovery Board put up two separate
websites in under six months, Recovery.gov and PaymentAccuracy.gov. “Government
usually takes years to do that,” he said. “We didn’t do it the way the
government usually does it but that’s probably what made it work.”

Challenges in data accuracy or clarity
in project updates
continue, however, despite Devaney’s
assertions. That said, there’s more to the story than data or website
infrastructure. As Jason Miller reported for Federal News Radio, the
Recovery
Board’s success inspires others
within government. It’s not the
success of the “RAT Board” in moving Recovery.gov to the cloud or redesigning the site: it’s the use of a data
visualization software for fraud detection.

This tool is a key a component of the enforcement of the Obama
administration’s “do not call” list
, which was announced in June
of this year.

Crowdsourcing and the power of open data

For those who weren’t familiar with the tool that Werfel or Devaney
described, the second panel that testified before the Senate provided
the answer: Palantir
Technologies
. Until TechCrunch’s recent coverage, in which Evylyn
Russell called Palantir
the next billion dollar company
, the data analysis software
developer was operating under the radar, at least outside of the
beltway.

Analysts within the intelligence agencies are using Palantir to
fight cybercrime. Transparency wonks are exploring Data.gov with AnalyzeThe.Us. And the Department of
Health and Human Services is using Palantir internally to detect fraud with the Medicare system. That innovation,
incidentally, is precisely why Palantir
is in the technology spotlight
at the Gov 2.0 Summit next month.

“We’re specialized in the least glamorized part of finding fraud,”
testified Alexander Karp, founder and CEO of Palantir. “Palantir is
based on a methodology that reduced fraud at PayPal from something
that takes thousand of hours to something that could be done in
real-time.”

When applied within government agencies and enterprises, Palantir
helps non-technical analysts see latent patterns in open data.
Effectively, it is a platform that allows subject-matter experts to
perform highly sophisticated analyses. “The inspector general
community has never had these tools before,” said Devaney in his
testimony.

Karp’s written testimony is embedded below:

Following Karp, Rob McEwen, founder and former chairman and CEO of Goldcorp, told a story about the power of crowdsourcing that will
be familiar to readers of “WIkinomics.”
As described in this excerpt,
Goldcorp published online every element of geographic data the company
held, investing nearly $1 million in prize money and website development.

Virtual prospectors spread to the site and, in time, identified more
than 100 sites on a 55,000-acre property that yielded 8 million
ounces of gold. “Incentives can be much more than cash,” said McEwen. “Nobody is as smart as everybody. The biggest goldmine in the world exists between everyone’s ears.”

After McEwen, Crane offered insights into MIT’s win in the
DARPA Network Challenge, as described above. I talked to Crane later
about crowdsourcing and government. Here’s our short interview:

Crane’s written testimony is embedded below:

Related:


The link between technical innovation and government improvement will be explored at the Gov 2.0 Summit, being held Sept. 7-8 in Washington, D.C. Request an invitation.

tags: , , , ,
  • Dave Atkins

    These examples of how technology can be useful may help incrementally advance government engagement with citizens, but I think the greatest change that needs to happen is a willingness to truly embrace the idea that openness and collaboration with “the public” is good. Accountability is good, but it is reactive and presupposes the old methods of governance would work–if only we could ferret out the waste, fraud and abuse.

    The real game-change will be when it becomes routine for policy to be discussed in the formative stages through public forums and collaborative tools like wikis. The revolutionary idea will be a shift from an accountability focus to an effectiveness focus–to stories of how crowdsourcing and collaboration are resulting in better policy and better programs–when it becomes unthinkable that an elected official would refuse to participate in a blog discussion or demand that participants identify themselves before responding. On the citizen side of course, that day will also need to be characterized by civil, rational, constructive engagement…

    These are great stories–but I hope the takeaway is that “engagement” is the value–not technology. We are not held back by a lack of technology–no more than we are held back by lack of intelligence. The great policy challenges are not to find answers but to navigate workable solutions through the politics of conflicting interests.

  • JZ

    As far as Tim O’Reilly’s concept of government, I do not like the idea that the analogy illustrates government as a *foundation* that all else is built on. O’Reilly has also said of gov doing the least and that enabling citizens to develop, etc. I prefer a clear mention that citizens be developers of the platform. Government workers are citizens are they not? They know people who know people and it’s all connected, not a separation of person and platforms/system. If we allow too much drift between person and state, we get closed data, closed process, less accountability, more corruption, more waste, less quality of life, etc.

    Related comments by many people are being posted in relation to open data for public transit at http://www.streetfilms.org/a-case-for-open-data-in-transit

    [ http://PhiBetaIota.net ]

  • Christina Morrison

    I would agree with Dave’s comment here – openness and public collaboration will be key to the continued improvement of our government. However, I think crowd sourcing and data mining tools can be used right now to make significant impacts in the government, and the tools themselves can pave the path for successful with open government. As Senator Carper mentioned in the testimony linked above, hundreds of millions of dollars could be saved by setting up a better system of counting the census. I think the potential cost savings afforded by new technologies like this could be a precursor to the embrace of more collaborative and innovative technologies in government.