Dear Wolfgang, thank you for that!
The interesting thing is, that features are bound to a specific object,
be it our planet. I think the issue deserves a guideline or an extension
of the scope note. I do not see a conflict to continue from P56 with
P46. The inference would be, that from feature "hair bundle Z" P46i
forms part of feature "head X" P56i is found on object "statue X"
follows "hair bundle Z" P56i is found on "statue X" with the same
quantification. P46 on its own would not introduce a quantification
constraint.
If we generalize P56 to E18 Physical Thing as domain, we loose the
uniqueness of the carrying object.
Best,
Martin
On 13/2/2013 6:36 μμ, Wolfgang Schmidle wrote:
Dear all,
This question has probably been asked before, but I couldn't find an answer:
How to model subfeatures?
In Arachne there are many data fields that describe parts of a statue. In general we do not attempt
to group these data fields, but probably it makes sense to group together at least the data fields
that describe the head or parts of the head. Currently we use P46 E22 for all parts of a statue.
For example, "Haarfrisur" ("hairstyle") is modelled as
E22 Man-Made Object (statue)
P46 is composed of E22 Man-Made Object (head)
P46 is composed of E22 Man-Made Object (hairstyle)
P62 depicts E90 Symbolic Object
P2 has type E55 Type (content of "Haarfrisur", for example "einzelne
Strähnen").
However, according to the scope note for E26 Physical Feature the head of a contiguous
marble statue should actually be modelled as a feature of the statue, i.e. "(statue)
P56 bears feature E25 Man-Made Feature (head)" rather than P46 E22, because one
cannot determine a natural border between the head and the rest of the statue. By the
same logic, the hairstyle is a feature of the head.
But E26 cannot have subfeatures, i.e. we cannot say "E22 (statue) P56 E25
(head) P56 E25 (hairstyle). Technically E25 (head) P46 E22 (hairstyle) would be
possible, but that doesn't seem to be the right solution. What would be the correct
way of doing it?
(Our current modelling avoids this problem, and since P56 is a subproperty of
P46 I guess it is good enough, even though it could be more precise. Would you
agree?)
Thanks,
Wolfgang
_______________________________________________
Crm-sig mailing list
[email protected]
http://lists.ics.forth.gr/mailman/listinfo/crm-sig
--
--------------------------------------------------------------
Dr. Martin Doerr | Vox:+30(2810)391625 |
Research Director | 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) |
|
N.Plastira 100, Vassilika Vouton, |
GR70013 Heraklion,Crete,Greece |
|
Web-site: http://www.ics.forth.gr/isl |
--------------------------------------------------------------
