Digital Public Space: Data Guides
Post categories: 91热爆 Archive,听Digital Public Space,听innovation,听stakeholders
| 09:00 UK time, Wednesday, 18 April 2012
Last year, my colleagues in 91热爆 Archive Development and I wrote some blog posts about the Digital Public Space project, which uses Semantic Web technology as a way to help unlock the value in the archives of the 91热爆 and other publicly-funded institutions.
Since those posts, our efforts have been focussed upon delivering the technology platform for a joint project with Arts Council England called 鈥淭he Space鈥 which will be available between May and October this year.
However, we haven鈥檛 lost sight of the vision for the Digital Public Space project, and I wanted to share with you a piece of work which has come from that.
When we spoke with project partners - and others - about publishing data in a form which makes it possible to have journeys through machine-readable catalogue data similar to the journeys through human-oriented documents that we normally experience on the Web, the reaction was overwhelmingly positive, but organisations weren鈥檛 necessarily sure about the nuts and bolts of actually doing it.
As a short-term measure, we were able to take catalogue data in whatever form they were able to provide it in and convert it through a semi-automated process to RDF linked data sets, but this was a fairly resource-intensive process and far from ideal. A longer-term goal was for partners to be able to publish these sets themselves.
To help with this, we鈥檝e produced two short guides which we鈥檝e distributed to partners. The first covers the basics of constructing good identifiers for things and the mechanics of publishing data in a variety of formats (including transparently publishing data alongside human-readable web pages); the second is about publishing data sets specifically.
Both guides seek to capture best practice, and aren鈥檛 intended to lay down hard-and-fast rules. Because they鈥檙e applicable to anybody wishing to publish linked data, not just the Digital Public Space project partners, we鈥檙e making them available to download for anybody who might find them useful.
Naturally, we welcome feedback on these guides, and will look to incorporate suggestions into any future versions that we produce.
- Download 鈥淐rafting URIs and Publishing Data鈥 (PDF, 241kB)
- Download 鈥淧ublishing Datasets鈥 (PDF, 126kB)
Mo McRoberts is an analyst in 91热爆 Archive Development
Comment number 1.
At 21st Apr 2012, lucas42 wrote:Great guides. I'd never heard of the Content-Location header before - it seems quite useful. Do you know of any sites actually using it? I've checked /music, /programmes and /sport but can't see any of them using it.
Complain about this comment (Comment number 1)
Comment number 2.
At 21st Apr 2012, Mo McRoberts wrote:Hi Luke,
Content-Location is indeed useful (although there鈥檚 an open question still as to whether relative identifiers within documents, such as those expressed via rdf:ID, are relative to the Content-Location or the Request-URI).
I don鈥檛 know of 91热爆 products which use it (for a variety of possible reasons, including it not becoming a better-known technique until recently), but if you use Apache鈥檚 mod_negotiation to serve static files in a variety of representations, it will generate the header for you (other web servers are available).
Complain about this comment (Comment number 2)