The British equivalent to Obama’s data.gov opened today. Over at ReadWriteWeb, Marshall Kirkpatrick points out the scale of the ambition involved:
At launch, Data.gov.uk has nearly 3,000 data sets available for developers to build mashups with. The U.S. site, Data.gov, has less than 1,000 data sets today.
[…][Unlike the US equivalent, the site] includes 22 military data sets at launch, including one called Suicide and Open Verdict Deaths in the U.K. Regular Armed Forces.
However, these are raw datasets. As Paul Clarke points out, the site only pays lip service to openness until someone comes along and turns these sets into useful reports and applications:
The only test of real success is: use. Not usefulness. Not theoretical use. Real use. Getting beyond the novelty application, the demonstrator, and the hobby lies at the heart of really untapping the potential of data.gov.uk.
Indeed, the figures that Techcrunch Europe report suggest that turning this data into something useful may be harder than it sounds:
So far over 2,400 developers have registered to test the site and provide feedback, [while] 10 applications have been created.
I left a comment on Paul Clarke’s post pointing out some potential pitfalls that may inhibit innovation, including the government’s insistence on licensing the data under Crown Copyright and their impartiality regarding Twitter. There’s also been some criticism around the lack of a common data format for each feed (although the RDF triple proudly displayed on the front page suggests this is likely to change).
Nonetheless, I believe this represents a huge step forward. Turning raw materials into useful, compelling applications that improve the users’ quality of life requires a huge amount of creativity and talent, and providing the data feeds in the first place is a crucial first step.
One response to “Open data at data.gov.uk”
Personally I don’t think that it’s a problem that they have just made the data available.
This is phase one and I’d much prefer them to dump the data in any old format and let people start hacking on it rather than wait until the interface/format/api is ready.
By all means work on this, but step one should be to drop the raw data.
Crown copyright is the next issue to deal with, but you can easily believe that using CC was the only way they could sell the idea of opening the data in the first place.