Constructing an open data platform

I spent the second afternoon of UK Gov Camp in the session on open data platforms for the public sector. I was fortunate enough to have a shared interest in this with Tim Davies who lead the group work we did. Tim has written up the development of a Charter of Engagement for open data that he and others did during the session on his blog.


Harry Harrold of Neon Tribe had also spoken to me about some ideas he had for prototyping what an open data platform should look like. He came armed with a variety of hi-tech gadgetry: he had cardboard, scissors, paper, sticky-backed plastic……I was a little disappointed not to see any empty washing-up liquid bottles, but you can’t have everything.

My interest in the session was in answering the question of how to construct a platform for publishing and reusing open data that best supports the people using it. So, a system?that both supports the people within organisations who need to publish data online and which also helps anybody who wants to make use of that data.

We spent the first hour discussing a variety of issues, whilst Harry started knocking up an idea of what a page on an open data platform might look like. We were lucky enough to have James Hendler with us who shared some of his experiences of helping to establish the US government’s Data.Gov site.

After the first hour we split into different groups.

I joined a group that was lead by Michelle Ide-Smith. She took us through describing different categories of users who might come to the site, why each of them might come to it and some scenarios describing how they might want to use it. Thanks to Douglas Guthrie for writing this up.

A Use Case Diagram for an OpenDataPlatform from #ukgc12

This was a really helpful way of looking at how a platform might be developed. It meant we spoke about how the prototype design that Harry had started could be amended and enhanced. It also meant that we started to consider the way it could help people to achieve their goals. The slideshow below shows the paper prototype that Harry developed.

Michael Grimes and I spoke about what the role of an organisation to stimulate some uses of the data it is publishing might be. We considered how to get people together to make use of data and how organisations might find out what people found useful.

We’d been talking about this just before the session wrapped up.

So, it was good to hear Tim talk about the 5 stars for open data engagement?that he,?Liz Stevenson,?Anthony Zacharzewski,?Jon Foster and?Jag Goraya had drafted

* Be demand driven

* * Provide context

* * * Support conversation

* * * * Build capacity & skills

* * * * * Collaborate with the community

I could immediately see that some of the quite loose conversation that Michael and I had is being codified in there.

It made me think of the Help Me Investigate: Education evening that I’d attended with Joy Gibara from Birmingham’s Looked After Children Education Service last December. There Joy was able to share a load of information about sources of freely available data relating to education.

It was both the value that people got from Joy’s knowledge and the pleasure that she obviously gained from sharing it that I thought was really useful and interesting about that evening. We need to encourage people who work with data every day to explain it to interested people in ways like this.

And that idea gels nicely with Tim’s comments under three stars – “Support Conversation”

Are there easy ways to contact the individual ?data owner? in your organisation to ask them questions about the data, or to get them to join the conversation? Are there offline opportunities to have conversations that involve your data?

I think we ended with a consensus around the table that public sector organisations are going to need platforms for publishing open data that are more than just websites for hosting datasets. And I think that after the afternoon’s work most of us came away with some ideas about what such a platform might consist of.

Thanks to everybody who came to the session.


  1. Very interesting, just found this. It fits in nicely with a discussion we were having about Hamburg’s planned open data portal (at the Open Government Stammtisch) last week. Because open data is a really broad field, it seems like a really good idea to have an inventory of which datasets are available, then publish the ones that can be published and list the ones that can’t. There will probably be a grey area where the public body will need a bit of pushing (even campaigning) to let go of the data, or bring it into a suitable state for publication, but enabling a conversation around an individual data set rather than just about all datasets in an undifferentiated way would be great. Even where there are legitimate barriers to publishing data, if someone really *really* wants that dataset, perhaps they can work out a way of getting round those barriers.

Leave a Reply

Your email address will not be published. Required fields are marked *