Should we explicitly recommend the use of codes for vocabulary values
over the terms?
For example, with Role, the MARC Relators define a code and term:
Code = act Term = Actor
Code = adp Term = Adapter
etc..
And AAT also define codes:
TermID = 25513 Term = Poets
TermID = 25515 Term = Screen Writers
etc..
The advantage is that we can support different languages
for the Terms. The disadvantage is that we loose the
human-readability for metadata instances.
Comments on this?
Also, along the same lines, we may wish to define
codes for our DCAT1 list. We may like to use the words
we now have as the "tokens" but allow more human and
language friendly names. (A bit like what we do with
the 15 elements.) So, we may have:
token = "person" name = "Personal Agent"
token = "org" name = "Organizational Agent"
etc..
Comments on this?
Cheers...Renato <http://purl.net/net/renato>
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|