Dan Matei wrote:
> -----Original Message-----
> From: Karen Coyle <lists_at_KCOYLE.NET>
> Date: Sun, 25 Oct 2009 10:21:10 -0700
>
>
>
>> Dan, I think these become relationships. There's no reason why the same
>> person can't be president of the US and governor of Arkansas -- those
>> are acts by that person, they don't change the "person-ness". But you
>> want both uses of Clinton to have the same identifier, because they were
>> both the same man, just in a different role.
>>
>
> Not good enough for me :-)
>
> Clinton with the president hat on (say, as author) is not exactly the same as Clinton himself or Clinton as governor.
>
> So, we should have:
>
> a (record for a) person: Clinton (himself, as a free individual :-);
> a (record for a) persona: the person Clinton associated with the legal body US;
> a (record for a) persona: the person Clinton associated with the legal body Arkansas;
>
> And those personas could be blank RDF nodes (RDF anonymous resources), but I prefer them to be URI-ed resources.
>
Sure, you can do it that way. It doesn't really matter what your
decisions are as long as they are clear and identified. You may wish to
link them to each other things through either relationships or
identifiers. So you could create a canonical identifier for Clinton
himself, and then the President Clinton can be a persona of that
Clinton. Or not, depending on your view. But ideally your data would be
designed in such a way that others could re-use it, and could even make
connections and add relationships of their own.
I'm not sure how we structure records at this point. To some extent I'm
avoiding thinking about it because the data I have worked with is so
record-based that I'm afraid that I can't think about it without bias.
It does get back to the question of FRBR entities and how we hang them
together while at the same time allowing re-use of the data outside of
the record that we have created. I *think* we do that by defining our
elements independently of the record, but I until I see it done (or do
it) I'm not sure that works.
kc
--
-----------------------------------
Karen Coyle / Digital Library Consultant
kcoyle@kcoyle.net http://www.kcoyle.net
ph.: 510-540-7596 skype: kcoylenet
fx.: 510-848-3913
mo.: 510-435-8234
------------------------------------
Received on Sun Oct 25 2009 - 16:14:45 EDT