[crm-sig] Mapping FRBR to CRM

martin martin at ics.forth.gr
Tue Feb 5 23:19:35 EET 2002

Dear Patrick,

Congratulations to this careful and thorough work. I try to work through the open problems
you have stated, but it is a lot to do.

So I think the first I would like to comment on is the mapping of Manifestation. Obviously
we have all problems with this concept. I think that the FRBR has not the logical rigidity
that the CRM uses to have. Therefore I propose to make a more free interpretation of this
concept. In a mapping exercise, when the definitions of concepts seem to be confusing,
one should resort to examples of use.

Some thoughts on Manifestation. If Manifestation is physical object, then what is an instance of it?
A book in my hand is an item. It is not an instance of manifestation.
Sets of things, which are as a total an instance, are necessarily instances of concepts (universals),
following all knowledge representation theories I am aware of.

Competing with this notion is the aggregate or collection of physical objects, which we have
(and others like Gerstl, Pribenow, Guarino) identified as a complex physical object. So may-be
the manifestation is a collection? In order to be, it must be maintained together, and be identified by
being kept together.

All that unites the items of a manifestation is the sharing of characteristics. The INTENDED
characteristics, those imposed by the plan of the production process. Under this consideration,
Manifestation for me is not material, but on the level of information objects, designs and
procedures (or metaclasses). Equal to the blueprints for a house, a car etc. The transition between general
characteristics and those we regard as precise, is often called "materialization".
In practical terms, the unique object, which carries an ISBN number, is not material.

The likeness, or "physicality" of a materialization is deceiving. The products look alike only
in the macroscopic view of people and under functional aspects. Deviations in chemical substance,
spots and impurities are normally overseen, as finger-print etc.

So, what makes a Manifestation? Is there anything I can recognize on an isolated sample, that
we have to do with a manifestation? I'd argue no. I think, what makes up a manifestation is, that
is the manifestation of an Expression. So I tend to see Manifestation as a specialization of
Information Object, which a) "is manifestation of" an Expression, and b) has characteristics
determining also the physical shape of its carriers.

The latter pushes Manifestation, and Expression, under the roles, i.e. classes which are defined
by relationship to other objects, rather than by the independent properties of the instances themselves.

What do you think?


PATRICK.LE-BOEUF at bnf.fr wrote:

> Dear all,
> Here is a first draft of a mapping FRBR/CRM. Please feel free to add any
> comments and remarks, or to ask any questions.
> Best wiwshes,
> Patrick
> (See attached file: Mapping_FRBR-CRM.doc)


 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/proj/isst         |

More information about the Crm-sig mailing list