I attended the “Creating a co-operative Open Data standard” event in Manchester, yesterday. Organized by Co-operatives UK, Open Data Manchester, and Co-operative News, the event brought together a range of people with helpfully diverse areas of expertise and motivation.
I found the talk from guest speaker Chris Taggart, of OpenCorporates, very interesting. OpenCorporates have gathered publicly-available data on companies around the world into an Open Data database, under the Open Database Licence. Chris explained that it’s difficult for individuals or smaller organizations to get a handle on the many companies that have subsidiary relationships within multi-nationals. The OpenCorporates database gives the ability to examine these relationships back to us.
Chris also explained that Governments are drowning in a sea of data that’s difficult to use because it’s stored in different, incompatible databases. data.gov.uk is a good example of the UK government starting to tackle that, by using Open Data standards, and even better, giving us access.
The Co-operative News are also promoting, and planning to provide, Open Data at thenews.coop/open.
Later, the group discussed the Why, What, and How of co-operative Open Data. Naturally, the part that grabs me is the technology, so I concentrated on “What does the technology look like?”
RDF immediately seems like an obvious choice for a format, perhaps also with Dublin Core metadata. For the techs amongst you, the RDF primer is a great place to learn RDF.
I posited, to Steven Flower of Open Data Manchester, the idea that we should use Tim Berners-Lee’s Linked Data concepts, rather than a central repository of Open Data. Different co-operatives could provide their Open Data in a way that it can be linked using the principles that Berners-Lee proposes.
All this reminds me of the concepts I saw in the EU-funded NEPOMUK social semantic desktop specification. I remember, when I tried out their Java proof-of-concept, the excitement of being able to link together disparate data easily, and quickly create definitions of new data classes. Their resulting ontologies are great examples of what can be done with RDF. Perhaps we can achieve something similarly exciting with co-operative data.
As lead developer of our Content Management System, I’ve also been led to think about how it might produce Open Data. As we add social-networking features, can we allow a site’s members to make parts of their profile accessible through Open Data and also give them the ability to control who (their friends, colleagues, etc) has access to which parts? Could we be part of a movement that facilitates the growth of social networks across multiple sites, rather than relying on a single site like Facebook?
The Co-operative News also have a good write-up of the event.
One Response to Creating a co-operative Open Data standard