Delicious
Digg
StumbleUpon
Facebook
Google
YahooToday the Obama Administration announced its new Open Government Directive, the product of months of deliberation, which orders agencies to take concrete steps to make public information available on-line in machine-readable open formats.
Before your eyes glaze over, think about what government data actually can tell us.
In "How Transparent is Your Turkey," I blogged about the many ways in which food safety information--data that we pay for as taxpayers and consumers--is not now available in useful formats. For example, let's say you want to find out if a particular brand of cranberry sauce has been subject to recall because of contamination. You can't, or at least not easily, because the FDA doesn't provide such information in a searchable database.
Today's New York Times investigation exposing that 20 percent of the nation's water treatment systems have violated key provisions of the Safe Drinking Water Act over the last five years comes from an analysis of public data obtained by the newspaper.
Programs already underway or planned by the government (for a list, see this blog post by my colleague Daniel Schuman) would do such things as open up public housing data so that it is easier for outside groups to evaluate and analyze where worst public housing is and recommend solutions.
Now agencies face a series of deadlines; for example, by Jan. 22, each agency must identify and publish online, in an open format "at least three high-value data sets and register these data sets via Data.gov." By February 6, they must create dedicated "open government" web pages; by April 7, each agency must develop its own open government plan.
At the Sunlight Foundation, where I'm a consultant, we'll be eagerly watching to see what happens next.
Nancy Watzman
consultant, Sunlight Foundation
www.muckrakingmom.com Because MUCK doesn't scare MOMs
I'm all for this, but I'm also thinking...
Kim Pearson December 9, 2009 - 1:37pm
One of the implications of having lots of public data online is that there needs to be a concomitant effort to ensure that the data is properly vetted and interpreted. As we know, any time you have a mountain of data, some of what you'll find will be good, but some of it will be riddled with errors. And particularly when one is looking at historical records, it's very easy to misread what one finds. I started thinking about a possible computational approach to this problem in a blog post I wrote yesterday. I'd be interested in your thoughts,
Beyond that, I'm excited about the possibilities this opens up to make this data more accessible. Thanks for writing about it!
Kim
BlogHer Contributing Editor|KimPearson.net|