The third Open Data Camp took place last weekend. I’m proud to claim the hat-trick of attending all three: the inaugural at Winchester in Feb 15, the second at Manchester in Oct 15, and this one at the Watershed, Bristol (great venue).

I’ve blogged before about how much I’ve enjoyed attending this type of event and OD Camp 3 did not disappoint.  There was an awful lot going on, with stacks of energy, enthusiasm, interest and intelligence in evidence.  The session grid lists 34 sessions and an @OwenBoswarva unkeynote.  There were skill swap opportunities and making of stuff (room temperature sensors publishing open data).  My  favourite ‘unsession’ was Ben Proctor trying his hand at telling stories with data: a must for fans of osprey love stories.

I started the first day sitting on my hands so as not to pitch for a session.  This was partly because I’d already pushed my ideas about knowledge management quite hard at previous events and partly because my Open Data & Supply Chain Management session had not gone down that well at OD Camp 2.  But it was mainly because I wanted to listen more – to get a better view of who was contributing and what themes were emerging.

The first session I attended was about engendering data literacy by nurturing an open data user/activist community of practice.  The session meandered a little, as can happen, and – like on an unplanned walk – we had the unexpected brought to our attention.  We enjoyed a philosophical diversion into the question of what data literacy means:  if it’s about understanding data and the context within which it is being used then isn’t that really a question of information literacy (if one defines ‘information’ as data in context)? We also touched on the problems of statistical illiteracy in the general population.

The session ended with a demonstration of the Data Campfire – a great idea by @natfoo and @wisewormm.  As a former Scout and Scout-leader, I found the image of a campfire, as a place to share and explore stories, instantly comprehensible.  As well as hosting stories about using data, the site allows users to discover and rate open data sources and also share data software/tools.  The over-riding sense of the importance of a community encompassing open-data providers, users and beneficiaries was strong.

It was then time for something a little more specific, with a session devoted to local open data.  I was interested to hear the sessions leads:  I’d met @MartinHowitt before, while the ‘Data Magpie’ persona  of @ldodds  had intrigued me for a while.  The discussion focussed on how to connect and motivate people in the collection, provision and exploitation of local open data.  The idea of organising meetings with local community service providers and activists to discuss their problems, before you organise a hack-day, seems obvious – once someone’s said it. I came away thinking there were four key things to consider: a clear purpose for the endeavour; a convincing WIIFM (what’s in it for me) message for individuals; making it easy to contribute data; making it just as easy for contributors to get data back out.

Having scaling down from national issues to the local, I was next offered a chance to go global with the session “Open data from space”.  This turned out to be a difficult topic to progress.  The starting point was ‘there’s a lot of open data available from satellites, but we (the open data community) don’t seem to be exploiting it’. We then ran through the reasons why that was the case with specialist knowledge and often specialist kit required.  My takeaway from this session was that a component of data literacy is understanding that some open data sets may be complicated and difficult to handle. Satellite data, gridded 4d numerical weather models, high-frequency sampled physical data streams all present difficulties in exploitation that a small CSV file does not.

Session Four brought a topic of particular interest to me – Linked Data.  Jen from Networked Planet had done a beginners’ tutorial on this topic at Open Data camp 2 and has also blogged four posts in a Linked data 101 series.  This session took things on a little further and gave @gklyne an opportunity to introduce Annalist – a linked data notebook.  This tool supports the collection, organization and sharing of structured and semi-structured data.  I look forward to exploring Annalist further – and will probably blog about it when I do.   In the midst of the discussion we had to pause and address the confusion between the term Linked Data as a specific concept in computing and the phrase linked data just as a description of two data sets that someone suggests are associated by causation or mere correlation. An important distinction.  I came away thinking that there is still a long way to go for Linked Data to fulfil its promise.

For my last session on Saturday I went to @alexrcoley‘s discussion on ‘Tips and tricks for finding a senior sponsor’.  This was a good exploration of DEFRA’s journey to the point of announcement by their SoS that within a year “…we will be making 8,000 datasets publicly available, in the biggest data giveaway that Britain has ever seen.”  Lots of good lessons about the art of stakeholder management, across semi-automonous agencies, progressive senior leaders and occasionally reluctant middle management. I recommend the session notes – they are well worth a read.

This post is turning out rather longer than I expected, so I’m going to publish ‘Day 1’ and then think about day 2.  As for the wrap-up, I think I’ll let @Drawnalism cover that.