• If you are citizen of an European Union member nation, you may not use this service unless you are at least 16 years old.

  • Buried in cloud files? We can help with Spring cleaning!

    Whether you use Dropbox, Drive, G-Suite, OneDrive, Gmail, Slack, Notion, or all of the above, Dokkio will organize your files for you. Try Dokkio (from the makers of PBworks) for free today.

  • Dokkio (from the makers of PBworks) was #2 on Product Hunt! Check out what people are saying by clicking here.


Permanent IDs

Page history last edited by Mia 12 years, 3 months ago

My question: should we look at a UK sector-wide solution for permanent URIs, should each institution mint our own, or should we look at some other, perhaps international solution?  And what can we learn from other sectors?  There's no definitive musicbrainz or wikipedia that provides an existing reference point.


An oldie but a goodie.  The importance of 'cool URIs' that don't change might be self-evident, but that doesn't mean it's understood by everyone in a cultural heritage institution.  Help us make the case for them here.


More interestingly, what happens if there's more than one URI for a single object? e.g. one in a museum in collection catalogue and a sector-wide list?


How do we get around the fact that museums can change their name, either through re-branding or structural change? And what happens when an object is deaccessioned, or content is created around a loaned object?


What other questions should we be asking?

Comments (8)

Mia said

at 4:25 pm on Apr 20, 2009

Bill Roberts said

at 2:08 pm on Apr 5, 2010

I think you should mint your own IDs. While a bit of careful thinking and design should go into this and it's a good idea to make an ID as 'permanent' as reasonably possible, the more important thing is to start taking advantage of the possibilities of Linked Data. With the recent efforts at data.gov.uk, there are best practices in ID design that you could follow as a starting point.

It doesn't matter if there is more than one URI for an object. Most things will ultimately have lots of identifiers, but it's easy to relate them so that users of the data (people or software) know if two identifiers point to the same thing.

For Linked Data to work best, you should make de-referenceable identifiers - ie if you stick the URI in your browser, then it comes back with useful information about that object. If an institution changes its web address, then you may lose the ability to dereference old identifiers. That's not ideal, but it's not the end of the world. The identifier itself is still valid and you can mint new identifiers in the new domain and make links between old and new.

While it makes sense to think through these issues, it's easy to get stuck in analysis paralysis and in general these types of design decision can be reviewed in the future.

Mia said

at 3:43 pm on Nov 3, 2010

There's some interesting discussion about this at the moment on various blogs:

Sebastian Heath: Progress on Museum URIs

Sebastian Heath: Responses to "Progress on Museum URIs"
(Key quote for me: "...the exclusive right of museums to identify their objects, that's a non-starter. It's neither the right thing to do nor is it possible. On first reading, I took his e-mail to represent a welcome assumption of responsibility by museum to provide a locus of stability for reference to their collections. But to be clear, objects will have multiple identifiers. Referring back a common identifier promoted by and discoverable at the holding institution will ease the process of recognizing that two or more identifiers refer to the "same thing". That will itself promote the idea of a discoverable and multi-vocal discussion about the past.")

Eric Kansa: Identifing Objects in Museum Collections

Eric raises the Elgin marbles as an example of a contested object where one particular institution minting a URL and thus 'claiming' to be the canonical home or representation of that object would be poorly received. Less controversially, but equally complex, which of the three museums that own part of Phar Lap (http://museumvictoria.com.au/pharlap/) - the hide, the heart, the skeleton - could claim ultimate authority for the Phar Lap record? (Assuming they wanted to do so for Phar Lap rather than just parts-of-Phar Lap).

Mia said

at 3:45 pm on Nov 3, 2010

Doh, forgot to add twitter handles: http://twitter.com/sebastianheath and http://twitter.com/ekansa and also http://twitter.com/leifuss (or @sebastianheath and @ekansa and @leifuss )

Mia said

at 1:17 am on Nov 4, 2010

And Bill - yes, you're right! It's just hard for museum people to cope with the concept of possible impermanence without feeling a bit faint.

One annoying quirk of Science Museum/NMSI accession numbers is that they contain '/' (used for parts of a whole) - I've never known what, if any, problems that might cause when building URIs. I discussed it a bit at http://museum-api.pbworks.com/w/page/Science-Museum-linked-data but need to go back and think over it and the comments again.

Mia said

at 1:31 am on Nov 4, 2010

I've been inspired to finish the Phar Lap post that's been in my head for years - 'What would Phar Lap do? AKA, what happens when Facebook and museum URIs meet a dead horse?' http://openobjects.blogspot.com/2010/11/what-would-phar-lap-do-aka-what-happens.html

Richard Light said

at 10:31 am on Nov 4, 2010

To me, a museum publishing a URI for each of its objects is simply the culmination of a process which begins when it assigns (internally) unique identities to those objects and continues when it publishes HTML descriptions for those objects on the web. It's a curatorial responsibility thing. Of itself, the act of publishing a URI doesn't make any claims about ownership (in any sense), though having minted the URI you can of course include assertions about who manages it, where it is, etc. And yes, other people can and will mint URIs for (some of) your objects - so what? Yours will be the only one which falls within the museum's domain (in the URL sense). People will be free to draw their own conclusions about which, if any, is "authoritative".

Phar Lap as a subject of discourse has, of course, already got a Wiki/dbpedia identifier (http://dbpedia.org/resource/Phar_Lap), inter alia. That shouldn't stop the three museums from publishing URIs for their bit of Phar Lap - in fact doing so allows you/them to say where the constituent parts of the horse now are.

Mia said

at 2:46 pm on Sep 10, 2012

Possibly useful: the Linking You Toolkit http://lncn.eu/toolkit/

"Linking You was a project at the University of Lincoln funded by the JISC under the 2011 Infrastructure for Education and Research Programme. Its aim was to look at and make recommendations for improving the way that identifiers for .ac.uk domains are planned and managed in higher education institutions.

The Linking You Toolkit is a collection of the outcomes of this project, and is intended as a starting point for institutions looking to better understand and manage their use of identifiers on their .ac.uk website."

You don't have permission to comment on this page.