Government Data Done Well, the EU Digital Agenda, and an unlikely unconference.

Last Monday, the European Commission held an unlikely unconference in Brussels, about “My big Idea for the Digital Agenda.” (In Brussels-speak, it was a “stakeholder day.”)

The idea: In order to help implementing Commissioner Kroes’ grand strategy for the Internet’s and Web’s future in Europe, crowdsource ideas, and let those who submitted them (and a few others) do the bake-off. Anybody could get into the room for this invitation-only event by submitting an idea on the Web. Ideas were refined and triaged in several rounds of ever growing groups (with some professional facilitators doing a fairly good job) during the morning, then presented to the plenary (and the Commissioner) in the afternoon.

Among the surviving ones: Beyond Raw Data: Public Sector Information, Done Well (with kudos to Jeni Tennison‘s talk at the ICT 2010 conference a few weeks ago).

The gist of the idea: Innovation based on public sector information will require massive data integration across diverse silos. Integration works best when there’s interoperability. Interoperability demands standards work. What standards, then, should be used for public sector information? And how can we forge agreements on what these are?

At this point, it’s worthwhile thinking a bit about the larger context, and about the directions these discussions could take.

In Europe, the Digital Agenda is emphasizing the importance of public sector information as an economic driver. One of the actions under the agenda is the commission’s review of the 2003 Public Sector Information Directive; the commission is seeking input till 30 November in a public consultation. That public consultation is asking a lot of good questions, for example around the cost for public sector information. (Imagine there’s a 1 euro price tag per data set…) That’s important.

At the same time, we’re finally seeing a lot of government data get out in the open. Some of it might just be in Excel sheets, some of it might be in documents, and some of it might even get four or five stars on the scale of linked, open government data.

As pointed out in Tim Berners-Lee’s Putting Government Data Online Design Issues paper and the W3C eGov IG’s Publishing Open Government Data Note, putting out some data is the first, important step in getting public sector information available, and opening it up for innovation and use by outside parties (and by those inside the government, too!). It’s a step that shouldn’t be waiting for the data inventory, the standards roadmap, and the standards development, all of which can take a lot of time.

But that doesn’t mean that raw data is enough, or that, with the raw data out there, everything will fall in place: Governments need to go all the way to the five stars, and we need to collectively — geeks and governments! — figure out the path to getting there. In the puzzle that the commission’s policy review (in Europe), the raw data movement, and the call to go all the way to five star data form, Monday’s idea adds the missing piece: Agreements about formats, agreements about vocabularies, agreements about how to put data online, and do it right.

Let’s start the discussion W3C’s eGov IG mailing list!