Data.gov Breaks 100k Mark (Well, Kinda)
BY Nancy Scola | Friday, July 24 2009
So blogged OMB director Peter Orszag this morning. Let's give him a chance to celebrate that remarkable growth, and then we'll, alas, have to quibble with a few details:
Today, Federal CIO Vivek Kundra announced that the number of datasets on Data.gov has increased from 47 to more than 100,000 – with new sets being added continuously.
Data.gov was created to open up the workings of government by allowing the public to easily access datasets generated by the Federal government to perform research, build applications, and conduct analyses. Since its launch in June, Data.gov has received more than 18 million hits.
Public participation and collaboration are critical components to the success of Data.gov. The public can suggest datasets they'd like to see, rate and comment on current datasets, and suggest ways to improve the site. The early response has been very positive. Individuals and organizations are not only viewing the data, but are also improving upon our work by analyzing and repurposing the information. The Sunlight Foundation recently launched Apps for America 2: The Data.gov Challenge to elicit from the public the most innovative applications based on the data available on Data.gov. One of the submissions is FlyOnTime.us. This site uses data from the Bureau of Transportation Statistics, available on Data.gov, to allow consumers to see estimated versus actual arrival times for flights on major commercial carriers.
As part of the movement towards transparency and openness, Vivek Kundra has been working with state and local governments to encourage them to open the warehouses of public data. A few have already taken up the challenge, and you can track their progress by clicking here.
Pulling together than 100,000 data sets (110,076, to be exact) in just a few short months is quite an achievement. It marks a giant leap over the few hundred feeds Data.gov has hosted until now, not to mention the status quo of the federal government in recent years. But it has to be said. When Orszag writes "the number of datasets on Data.gov has increased from 47 to more than 100,000," it leaves an impression that isn't entirely accurate.
What Data.gov is doing with today's growth is acting as a clearinghouse for data sets maintained by other government agencies, like EPA, Commerce Department, Department of the Interior. Those data sets can be notoriously unwieldy, broken, or otherwise difficult to work with. That's part of the dream of Data.gov -- that we might have one central place to go where the data is so well done that it increases the odds that citizens, the media, and advocacy groups will make use of it.
Click around the new data sets listed under "GeoData" on Data.gov, though, and you'll notice that will some are locally-hosted GIS data files, many aren't actually "on Data.gov" at all. They're hosted by other offices, research institutes, and agencies. Click on the entry for a zooplankton biomass and species survey, and you're pushed off Data.gov to a site hosted by a consortium of public and private researchers.
In its own way entirely sensible. Keeping 100,000 data sets current and well crafted is probably a Sisyphean task for an agency the size of OMB. But there are challenges with acting as a clearinghouse rather than a data host. Click on the entry for a North American butterfly survey, for example, lands you on a U.S. Geological Survey page which reads "Butterflies & Moths of North America have moved!" And a Data.gov entry for a crop modeling study points you to a NASA.gov page that requires registration before you can download the data set.
Data.gov's growth has demonstrated the commitment of the Obama Administration to freeing government data across the full range of government. Deserved kudos there. But we haven't arrived at the promised land yet.