For a few months now we have been working on a new tool that we hope will help improve the way in which people publish their open data. Today we’re proud to announce the Open Data Certificates website beta.
The certificates, which are generated for free, give data publishers a way to:
There are three things that I wanted to highlight.
First, if you have published open data at all you’ve already achieved something. Even publishing a spreadsheet under an open licence can mean making a case to Management, battling with Legal and cajoling IT. At its most basic — Raw level — an Open Data Certificate shows you are publishing accessible data under an open licence.
Second, the Open Data Certificate looks at the whole context of data publishing, not just licences and data formats. In framing the questions and working out the levels, we have taken the perspective of someone who wants to use the data you are publishing.
“What would it take,” we ask, “to invest time, energy and money in building something with this data?”
The answer is, “quite a lot”. Raw data releases in poor formats with few guarantees and little support are fine for experimenting with during hackdays, but entrepreneurs and existing organisations need a more reliable data supply. That doesn’t mean raw data is bad — far from it, it’s through these experimental releases that publishers can learn what people need from them — only that it is the start of a journey, not an end.
Third, the highest bar — the Expert level of the certificate — is really (no, really) high. The Expert level is what we expect from data that forms the core information infrastructure for a country. It marks a level of reliability, quality of provision and support that is important for data that underpins multitudes of services and decisions.
We don’t know of any services that could gain an Expert certificate at the moment. We hope that the certificate will push the open data services that are already good to address the areas they’ve been neglecting. We don’t know which will be the first to achieve an Expert certificate, but it will be a cause for celebration when they do.
But not all open data needs to be that good: publishers need to choose for themselves whether and where to invest resources to improve how they make open data available.
To finish, I wanted to thank all those who have contributed to the development of the Open Data Certificates thus far: all the people who have raised issues on GitHub, emailed with questions and suggestions, helped with our user testing, and offered support. In particular, many heartfelt thanks to Adrian Phillpot for his designs, Chris Kenworthy for his work on the copy, and the team at White October for developing the site.