PubMed Central ready for the Semantic Web and Open Data?
With a background in archaeology, the collections of the Archaeology Data Service tend to be rather more intelligible to me than the underpinnings of advancement in the life sciences that are the bread and butter of PubMed Central and its UK manifestation. I don’t need to be able to engage at a detailed level with all of the science to have long been able to see the importance and value of the initiative, or to see it as an important plank in wider open access and open data activities that I do understand.
As such, it was intriguing to return to decent network connectivity and pick up Matt Day’s post to Nature’s Nascent blog earlier this week;
“What do UK researchers want from UKPMC, their free archive of full-text journal articles? That was the subject of a meeting today at the Wellcome Trust in London.”
Matt mentions that UKPMC are surveying anyone with something to say, asking that we,
“please help [UKPMC] create the definitive portal for biomedical and health research literature”
The online survey is only available until the end of this week, but anyone adding contact details to their submission will be entered into a prize draw to win a weekend in London. There’s no small print, so I can’t tell you whether that prize will include the return flights from Auckland, San Francisco or Beijing… but I suspect not.
We’ve talked about open access publishing, and the publishing-specific requirements of open data, here before. Peter Murray-Rust’s podcast from last year is a good introduction to the area, and it would be a significant step ‘just’ to have some of his quite reasonable requests implemented more widely. Since that podcast, of course, we do now have a licensing framework that has scientific research as one of its intended use cases. UKPMC, the Wellcome, JISC, and others should certainly be looking at that, and giving far more serious thought to the ways in which the core data behind scientific advancement can be at least as addressable as the academic papers they’re investing so heavily in making openly accessible.
Returning to the survey, and Matt’s notes from the meeting, I was interested to see frequent mention of ‘Web 2.0 community participation;’
“Discussion groups then pondered the question about priorities for further UKPMC development. Three broad areas were put forward to focus us all: adding in Web 2.0 community participation, adding in new content types, and adding in new user services.”
Participation is certainly an important step forward, especially in a process that is as participative as science. My immediate impression, though (and I admit that I haven’t seen Timo Hannay’s slides yet) is that I’d question the real value to science or scientists of ‘community participation’ inside yet another silo.
We’re very good at carving the world up, and expecting people to interact in very different ways across the little pockets of functionality that each new web site offers. Yet another username and password. Yet another user interface. Yet another set of ‘friends’ to find, invite, and nurture. Yet another great site that you don’t use enough, because it’s too much hassle to recreate yourself and your network there.
Rather than ’simply’ bolt some tags, curvy-edged boxes, and other Web 2.0 frippery onto the outside of a repository, why not think more radically? Why not throw the whole thing open, leverage the power of the Semantic Web that solutions like the Talis Platform offer, and see what we can do to maximise the value of those connections between papers, people, and areas of research… rather than simply expecting people to restate them, yet again, for some half-purpose?
Most of the data already exist. The technology already exists. Let’s join the pieces together, and do something really interesting.
And, in related news, it’s great to see Richard Poynder writing in ComputerWeekly about some of the work Peter Murray-Rust has been putting in on open access and open data.
Technorati Tags: Nascent, open access, open data, Open Data Commons, Nature, Semantic Web, Talis, UK PubMed Central, UKPMC, Web 2.0, Web 3.0, web services















February 7th, 2008 at 11:04 am
Hello Paul, I’m interested in what you think makes the Talis platform (and the semantic web) different to all the other “silos” that are on offer? What do you suppose they offer that no other content management systems do? Is Talis Open Source software like DSpace for example, as this is often an important requirement for scientists?
As far as Open Access goes, Peter Murray-Rust is a good start but you might also find the following (Open Access) papers useful background if you are interested.
Cheers, Duncan
Matthew Cockerill (2005) BMC bioinformatics comes of age. BMC Bioinformatics, 6
Gunter Eysenbach (2006) Citation advantage of open access articles. PLoS Biology, 4(5)
Peter Suber (2002) Open access to the scientific journal literature. Journal of Biology, 1(1)
February 7th, 2008 at 12:04 pm
Hi Duncan, and thanks for the links.
I could answer at length, but maybe a comment’s not the best place for that. A big difference, I think, is that we’re actually able to move beyond the silo. Open Data, exposed using Open specifications like RDF, SPARQL, etc is visible and searchable across the web. Yes, the data has to sit somewhere (in a Talis Platform store, for example), but it’s available for combination with other resources elsewhere in the Platform or out on the open web. The data’s portable, too… so you can take your data out of the Platform and transfer it somewhere else if required.
The Talis Platform isn’t a straight comparison with a document repository like DSpace. Instead, it’s a set of services and capabilities that make it far easier for ourselves and third party developers to produce applications that will benefit from the ability to search, manage and navigate connections between disparate resources.
If you’re interested, we could set things up to let you load up data of your own, to see what it - and you - can do…
February 7th, 2008 at 3:15 pm
Quite by coincidence, in a meeting just this morning I suggested that one could envisage delivering the functionality of the systems we tend to call “repositories” in a “Webby” way using something like the Talis Platform.
I admit I found it slightly dispiriting to look at the diagram here
http://www.mkbergman.com/?p=416
and realise that - with one or two notable exceptions - so little of our “repository” metadata is, currently at least, visible in that landscape.
February 7th, 2008 at 3:17 pm
Pete - exactly!
Want to try?