Dear all,
Martin and I have had a very short discussion about how to model
information from old and not properly written catalogues from the
entographical museum in Oslo:
"We are working on a database for ethnography (the museum) and have as an
objective to make this database as CRM-compatible as possible. As usual the
real data are XML-tagged aquisition catalogues (19). In some cases one will
find entries simplely describing a mask very briefly and then at the bottom
its written Borneo. According to the model thought out by the ethnographers
this could mean: made or used (by people) or collected/bought (by
ethnographers) on Borneo. According to the model thought out by the
ethnographers this could mean: made or used (by people) or collected/bought
(by ethnographers) on Borneo. [Allthough if pressed they tend to deny
everything except that there seems to be a connection]."
Due to the really unclear connection it is tempting (at least for my
collegues and me) to model this by introducing an empty property P0 and
using a E13, attribute assignment to represent the fact that the connection
is documented in a certain (identifiable) catalogue. I got the following
answer from Martin:
"Physical Object - was present at - Event - took place at Place.
I don't think you need more. You could even classify the event with
a "MAKING USE or CoLLECT" Type. Totally precise. "
Martin is completely right that it iss totally precise. Since then we have
discussed the problem with Stephen and have been informed that we are not
the first to suggest a P0. A P0 will be useful from a information-logical
point of view. However, I see the point that the introduction of such a
thing will make it too tempting to collaps a mapping in to CRM into P0s.
Thus Martin's solution should be selected. The use of a "MAKING USE or
CoLLECT"-type will for many persons be a little unfamiliar. But in a proper
thesaurus it is fine.
Praeterea censeo E3 Condition State esse delendam
Christian-Emil
X-From_: [email protected] Fri Feb 06 16:39:10 2004
Envelope-to: [email protected]
Date: Fri, 06 Feb 2004 17:39:02 +0200
From: martin <[email protected]>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:0.9.4.1)
Gecko/20020508 Netscape6/6.2.3
X-Accept-Language: en-us
To: Christian-Emil Ore <[email protected]>
Subject: Re: CRM cross-reference Manual
X-MailScanner-Information: This message has been scanned for viruses/spam.
Contact [email protected] if you have questions about this scanning
X-UiO-MailScanner: No virus found
X-UiO-Spam-info: not spam, SpamAssassin (score=0, required 12)
Dear Christian-Emil,
What about this:
Physical Object - was present at - Event - took place at Place.
I don't think you need more. You could even classify the event with
a "MAKING USE or CoLLECT" Type. Totally precise. If you like this, please
post the question and solution to the crm-sig list. Others might be
interested in it.
BTW, do still think about a CRM meeting in the Norwegian mountains??
Cheers,
Martin
Christian-Emil Ore wrote:
Dear Marin
I got the material and have started on the job.
By the way: We are working on a database for ethnography (the museum) and
have as an objective to make this database as CRM-compatible as possible.
As usual the real data are XML-tagged aquisition catalogues (19). In some
cases one will find entries simplely describing a mask very briefly and
then at the bottom its written Borneo. According to the model thought out
by the ethnographers this could mean: made or used (by people) or
collected/bought (by ethnographers) on Borneo. The three cases will
clearly create different activities in a CRM-database. A simple way to
model this is by using E13 Attribute assignment and P141, P140 and
perhaps P70. This can be slightly unecessary or complicated. It is
tempting to extend CRM with a P0 being the superproperty of all
properties. This will make the model more "complete" in the sence that is
possible to abstract/generalise/every property as fare the left as one
may wish. Today connections between two or more objects may disappear if
one abstract to much. The above Borneo-example shows (in my opinion) that
one may find very underspecified relations in real data.
Christian-Emil
At 14:19 29.01.2004 +0200, you wrote:
Please check!
Best,
Martin
--
--------------------------------------------------------------
Dr. Martin Doerr | Vox:+30(2810)391625 |
Principle Researcher | Fax:+30(2810)391638 |
| Email: [email protected] |
|
Center for Cultural Informatics |
Information Systems Laboratory |
Institute of Computer Science |
Foundation for Research and Technology - Hellas (FORTH) |
|
Vassilika Vouton,P.O.Box1385,GR71110 Heraklion,Crete,Greece |
|
Web-site: http://www.ics.forth.gr/isl |
--------------------------------------------------------------
--
--------------------------------------------------------------
Dr. Martin Doerr | Vox:+30(2810)391625 |
Principle Researcher | Fax:+30(2810)391638 |
| Email: [email protected] |
|
Center for Cultural Informatics |
Information Systems Laboratory |
Institute of Computer Science |
Foundation for Research and Technology - Hellas (FORTH) |
|
Vassilika Vouton,P.O.Box1385,GR71110 Heraklion,Crete,Greece |
|
Web-site: http://www.ics.forth.gr/isl |
--------------------------------------------------------------
