On Sun, Nov 1, 2009 at 22:48, Dan Matei <Dan_at_cimec.ro> wrote:
> NB. I'm concerned here only with "subject identifiers" (not with "resource identifiers/locators").
Karen asks if these two aren't the same, or say that the difference
doesn't matter. My answer is, well, they are very distinct things
which happens to correlate in the way they look. One is a URI that
resolves, the other is just a piece of string of characters, so they
have two distinctly different semantics.
> Case 1.
...
> How to identify conveniently the Habsburg Empire's territory ?
Geographic territories are inherently difficult to mark up in a
non-graphic way. I can say "King Harold was the King of Norway" and
refer to a specific period in time, yet the geographical markup of
that time was substantially different from the the times before and
after. In the RDF world, I have no idea how to best do it. I think a
stream of GPS coordinates and the such is massively overkill, but
perhaps the only thing that will be precise enough? (Well, precise
until you discover that borders are a very imprecise science of the
past)
In a Topic Map we often have topics that represent the whole, so ;
http://library.org/history/europe/power/Habsburg_Monarchy
If you want to be specific about certain markups within it, they are
in essence their own subjects ;
http://library.org/history/europe/power/Habsburg_Monarchy/Beginnings_(1278%961526)
http://library.org/history/europe/power/Habsburg_Monarchy/Austria_and_The_Reformation_(1526%961618)
http://library.org/history/europe/power/Habsburg_Monarchy/Austria_and_The_Thirty_Years_War_(1618%961648)
http://library.org/history/europe/power/Habsburg_Monarchy/Austria's_Rise_to_Power_(1657–1714)
http://library.org/history/europe/power/Habsburg_Monarchy/Charles_VI_and_Maria_Theresa_(1711%961780)
http://library.org/history/europe/power/Habsburg_Monarchy/The_Reigns_of_Joseph_II_and_Leopold_II_(1780%961792)
http://library.org/history/europe/power/Habsburg_Monarchy/The_Era_of_the_French_Revolution_and_Napoleon_(1792%961814)
http://library.org/history/europe/power/Habsburg_Monarchy/The_Nineteenth_Century_(1815%961918)
Specific topics can either be associated with the whole, the part, or both.
> The case applies also to the birth/death places of persons in authority files, no ?
Yup.
> In our movable heritage database, we have - of course - to record the "is keeper of" in CRM terms) relationship between
> the object and the museum. But, for instance, the local authorities of my home town periodically change the
> organisation of the city museum. So from time to time, the art museum, the history museum and the ethnographic museum
> are merged in one museum, and after a couple of years, they are made autonomous again. Of course, always for a better
> management :-)
>
> So we keep identifiers for 4 entities and try to keep track of the time spans for each.
Well, if identity management was done in a distributed fashion, you
create identifiers for all of these bits, and use relationships to
describe their current markup. Distribution should work out the right
identifier at any given time, and this process is pretty much just
like what catalogers are already doing with copy cataloging (except
that the tools will be slightly more funky).
> The same case applies to the many reorganizations of our brave government. The ministries are combined and recombined
> many times. (Is that a Romanian peculiarity ? :-) What to identify ? Only the legal entities (the ministries) or also
> the "functions", say "central cultural administration" which sometimes is performed by the Ministry of Culture and
> sometimes by the Ministry of Education and Culture ?
The USSR still exists as a subject even if the entity doesn't exist
anymore. If you want to say that a report was written by the
Department of Silly Walks, then an identifier for that, even if the
Department Silly Walks later was merged with the Office for Funny
Laughs into the Ministry for Stupid. Easy of them really deserves
their own identifier; how else are you to talk about them, even if in
abstract, non-existing ways?
> We speak about:
> a) Wittgenstein I and Wittgenstein II;
> b) Marx and Young Marx;
Again, the same;
http://library.org/people/Wittgenstein
http://library.org/people/Wittgenstein/early_years
http://library.org/people/Wittgenstein/later_years
http://library.org/people/Wittgenstein/later_years/determinition
http://library.org/people/Wittgenstein/later_years/epistemology
http://library.org/people/Wittgenstein/death
Be specific when you need to be specific. Oh, and don't let the cool
URI's fool you; the taxonomy inherit in the URI's are not to mean that
you should create taxonomies for this, even if you can. Our two URI's
for Wittgenstein might just as well be ;
http://library.org/id/34938490348
http://library.org/id/3435345340348
Define their relationships so that query engines can parse it, though.
That's the way to go.
Regards,
Alex
--
Project Wrangler, SOA, Information Alchemist, UX, RESTafarian, Topic Maps
--- http://shelter.nu/blog/ ----------------------------------------------
------------------ http://www.google.com/profiles/alexander.johannesen ---
Received on Sun Nov 01 2009 - 21:27:44 EST