[crm-sig] Modeling names and pseuds of actors

martin martin at ics.forth.gr
Wed Mar 20 15:58:51 EET 2002


Dear Patrick,

Thank you for the interesting reference.

I think:

"ALEG follows the Indecs model in which works are linked to names, and then, 
in a separate step, names are linked to agents." 

 is a pretty bad model for museums. E.g. the Master of the Paradise Garden
is known by work, without a name. Where a work is, there must be a person. There
need not be a name.
One name may point to different persons, so a link first to name, then to
person would connect all persons with that name to that work. One should introduce
unknown persons instead, concluded by evidence = by their work. This may produce
more persons than there existed.

I think "pseudoperson" is a nice concept. For the CRM I would argue, that what 
matters to register an Actor is not so much its existence, but the documentalist's
believe in it. In other words, I would see Person as a superclass of real and 
pseudopersons. We may at a first stage believe both are real. Then we learn, that
initially two persons are one. This happens frequently with all kinds of
"Existence" (Persistent Item) when increasing or merging knowledge. One could use
a link : "identical with" to denote the fact, before merging both, or when a real
"pseudoexistence" is maintained. Having one thing split into two does not
lead to wrong conclusions, only to less conclusions than if we knew it were one.
On the other side, having two things be taken as one leads definitely to wrong conclusions.

To bind a person to a unique "most common writing name" can be seen as an identifier
for the person. In the proper sense of the word, this is no good choice for historical
data, as people will change ("This records what people would expect to see as the name of that Agent").
Different nations and future nations will have different practice.



best,

martin

PATRICK.LE-BOEUF at bnf.fr wrote:
> 
> Dear all,
> I've just found, on the AustLit Gateway site, a page about the way they modelled
> names and pseuds of agents. I think it may be interesting, with respect to our
> recent discussions in Monterey. Here is the URL :
> http://www.austlit.edu.au:7777/DataModel/relationshipPseuds.html
> Best wishes,
> Patrick

-- 

--------------------------------------------------------------
 Dr. Martin Doerr              |  Vox:+30(810)391625         |
 Principle Researcher          |  Fax:+30(810)391609         |
 Project Leader SIS            |  Email: martin at ics.forth.gr |
                                                             |
               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               |
--------------------------------------------------------------



More information about the Crm-sig mailing list