A little less than a month ago the Open Data Institute launched Open Data Certificates, a new scheme to promote open datasets and help build trust between data producers and data consumers. Essentially, data producers fill in a detailed online survey for a particular dataset, covering various topics such as legal information (rights, licensing, privacy), practical information (findability, accuracy, quality, guarantees), technical information (locations, formats, trust) and social information (documentation, support, services). Based on the answers, a certificate is created (see this example), with one of four different ratings, from Raw, over Pilot and Standard to Expert (where Expert is an exceptionally high rating that so far no one has achieved). The certificate comes complete with code that the data producer can embed on their dataset’s webpage (just a badge or the complete certificate), as a way to inform and reinsure potential users of the quality of the dataset.
Certificates are “self certified” (and say so at the top). In other words, the ODI does not verify or review the submitted forms in any way. However, the terms state that the ODI reserves “the right to revoke a open data certificate at any time”. So, if future Open Data fraudsters (I wonder when we will first see such a thing) lie when filling in the form, in order to get an Expert rating, the ODI can just remove the certificate. This is possible, because the certificate always remain on the ODI’s website; the data owner only links to it.
In a way these certificates are similar to Tim Berners Lee’s 5-Star Open Data scheme. However, where the 5-Star scheme focusses mainly on technical aspects such as data formats and licences, the OD certificates rate datasets on a much wider basis (where technical aspects are just one component), covering “soft” aspects such as ease of use, timeliness, documentation, support, reliability, etc. Also, the well-designed and good-looking tools (forms, badges, copy&paste code) available on the certificates website makes them much easier for data producers to use. Of course, that they come from a respectable organisation like the ODI, gives the certificates additional weight – both for consumers, who will trust them more, and for producers, who will be keen to have them.
The ODI’s Gavin Starks calls the new scheme the “first robust quality badge for open data”. I believe this is true, as far as this is possible for a self-certified badge. What I particularly like is the fact that we now have something like the counter part of a licence: where a licence dictates how a consumer can use a dataset, the certificate makes promises and assurances about quality from the side of the data producer. We’ll see how it works out, but I have been thinking for a while that we really need something like this for Open Data to see broader use.