Archive for the ‘well’ Category

An elastic bucket down the data well (#rdtf in Manchester)

Posted on April 20th, 2011 by Paul Stainthorp

I was in Manchester on Monday for Opening Data – Opening Doors, a one-day “advocacy workshop” hosted by JISC and RLUK under their Resource Discovery Taskforce (#rdtf) programme. I delivered a five-minute ‘personal pitch’ about Jerome, open data, and the rapid-development ethos that’s developing at Lincoln.

Ken Chad is writing up a report from the day and Helen Harrop is producing a blog, both of which will be signposted from the website: http://rdtf.mimas.ac.uk/

The big data question

All the presentations can be viewed on slideshare, but there were some particular moments that I think are worth picking out:

The JISC deputy, Prof. David Baker was first up. His presentation, ‘A Vision for Resource Discovery‘ should be compulsory reading for university librarians. See, in particular, slides #6 (guiding principles of the RDTF), #8 (a future state of the art by 2012), and #11 (key themes).

Slide from David Baker's presentation Slide from David Baker's presentation Slide from David Baker's presentation

Following this introduction, there were three ‘perspectives’, short presentations “reflecting on the real world motivations and efforts involved in opening up bibliographic, archival and museums data to the wider world”: from the National Maritime Museum, the National Archives

…and from Ed Chamberlain of (Jerome’s ‘sister project‘) COMET (Cambridge Open METadata), the perspective from Cambridge University Library on opening up access to their non-inconsiderable bibliographic data. N.B. slides #4 (what does COMET entail?), #9 (licensing) and—more than anything else—slide #16 (“beyond bibliography”).

Slide from Ed Chamberlain's presentation Slide from Ed Chamberlain's presentation Slide from Ed Chamberlain's presentation

The first breakout/discussion session which I sat in on looked at technical and licencing constraints to opening up access to [bib] data. This was the point at which the tortured business metaphors started to pile up. ‘Buckets’ of data. ‘Elastic’ buckets that can expand to include any kind of data. And (my personal contribution, continuing the wet theme): data often exist at the bottom of a ‘well’. Just because a well is open at the top, it doesn’t necessarily make it easy to get the water out! You need another kind of bucket – a service bucket that makes it possible to extract and make use of the water. Sorry, data. What were we talking about again?

Then a series of 5-minute ‘personal pitches’, including mine just after lunch. I didn’t use slides, but I’m typing up my handwritten notes on Google Docs and I’ll post them as a separate blog post when I get a chance.

David Kay (SERO), Paul Miller (Cloud of Data) and Owen Stephens delivered the meat of the afternoon session in their presentation, ‘The Open Bibliographic Data Guide – Preparing to eat the elephant‘. The website containing the Open Bib Data Guide (which has not been formally launched until now) can be found at: http://obd.jisc.ac.uk/

The site itself is going to be invaluable in hand-holding and guiding institutions through the possibilities in opening up access to their own bibliographic data (OBD). Slides from the presentation that are particularly worth noting are #8 (which shows the colour-coding used to distinguish the different OBD use-cases) and #14 (examples of existing OBD).

Slide from the OBD presentation Slide from the OBD presentation

Paul Walk’s presentation, ‘Technical standards & the RDTF Vision: some considerations‘, is the source of the slide which I photographed (at the top of this blog post). Paul talked about ‘safe bets’; aspects of the Web that we can rely on playing a part in allowing us to create a distributed environment for resource discovery: including “ROASOADOA” (Resource- / Service- / Data-Oriented Architecture), persistent identifiers, and a RESTful approach. See also this blog post.

In the second breakout/discussion session, we discussed technical approaches. One of the themes which we kept coming back to was that of two approaches (encapsulated by Paul’s slide) which—while not mutually exclusive—may require different business cases or different explanations in order to be taken up by institutions. We characterised the two approaches as:

  • Raw open data vs Data services
  • Triple store vs RESTful APIs
  • Jerome vs COMET (bit of a caricature, this one, but not entirely unjustified!)

I was gratified that Lincoln’s approach to rapid development and provision of open services was also referred to in non-ungratifying terms, as a model which could be valuable for the HE sector as a whole.

Finally, we heard what’s next for the #rdtf programme. It’s going to be rebranded as ‘Discovery‘ and formally re-launched under the new name at another event: ‘Discovery – building a UK metadata ecology‘ on Thursday, 26 May 2011, in London. See you there?

Ken Chad is writing up a report from the day and Helen Harrop is producing a blog, both of which will be signposted from the website: http://rdtf.mimas.ac.uk.