[Crm-sig] Identifiers for activities about objects
joshan.mahmud at gmail.com
Thu Jan 17 19:54:29 EET 2013
We have had a similar issue with British Museum data and came up with the
1. Every object in our collection has this URI: *
is the unique identifier for the object.
2. For each object, there may be different areas we wish to create URIs
for such as different sides of a coin. For these we create:
3. For events; we have things like production, acquisition etc. and for
each one there may be multiple - i.e. multiple production events to a
single object as it was made by different people at different times. Or
multiple conservation events to an object as it was conserved many time.
For these we just use a simple counter and increment each one:
Each one of these will be an event (will have an rdf:type to a CRM class
which will be some sort of event) and each one will be detailed with who,
when. why etc.. that particular event occured.
The reason it has been done like this is because we do not have a unique
identifier for the event in our database - nothing to symbolise it to help
us generate the URI uniquely therefore it have been generated in the
context of that particular object - this may not be the case with your
For example take these two events:
Are two separate production events to two different objects - but actually
they were the same event - since the person made them at the same time for
example. In this case one could at some point assert that they are the
same event and therefore you can use owl:sameAs.
Hope this helps!
On Thu, Jan 17, 2013 at 3:44 PM, Athanasios Velios <a.velios at gmail.com>wrote:
> We are working on the detailed description of museum objects for
> conservation. I was wondering whether there are any guidelines on the
> production of identifiers for:
> 1) Object components
> 2) Activities on objects
> Complex objects could have hundreds of components, each requiring an
> identifier. They could also have a long history of modifications with a
> variety of materials and techniques, all of which again need an
> identifier. A unique number generated from a UUID tool, or similar, may
> work, but how does a human make sense of the seemingly random strings?
> Is there a generic recommendation for something that would suit humans
> in addition to computers?
> Dr. Athanasios Velios
> University of the Arts London
> Crm-sig mailing list
> Crm-sig at ics.forth.gr
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Crm-sig