I agree fully with John's assessment (criteria) and would actually go
further in defining a concept as nothing but a (numbered) placeholder. It is
meaningless without relations. In my system (NeuroCollective) a relation can
be semantic (creating semantic networks) but all concepts allow NL relations
which go into syntactic representations. In the very simplest form I have: (01)
Concept (X)
Concept (L1)
Concept (L2) (02)
Then triplet relations: (03)
(target, relation, value) (04)
preload the system: (05)
(L1, L1, "is called in English") // defining L1 as English language
(L2, L2, "is called in German") // defining L2 as German language (06)
now I can define: (07)
(X, L1, "table")
(X, L2, "Tisch") (08)
In reality the value field actually goes into a set of syntactical forms
that still represent the same concept (plurality, syntactic role, case,
conjugation.) play a role.
I can request X in various textual forms: (09)
PresentAs(X, L1, NP, plural, ...): tables
PresentAs(X, L2, NP, plural, ...): Tische (010)
The method PresentAs() returns also the correct article and IPA, to be used
if necessary. (011)
The NeuroCollective allows identifying X by either "table", "tables",
"Tisch" or "Tische", and some more forms in German as well as forms in any
other language and script that was learned. (012)
This is all hopefully an old hat. (013)
It gets interesting when the concept is complex. Here a new observation
applies that I just recently defined as some kind of eternal rule. (014)
The meaning of a concept cannot be truly defined until it can be translated
(e.g. into another language, or circumscribed in the same language) and n >
1 parties agree with the translation. (015)
In other words: A single language ontology may, but likely is never
correctly interpreted the same way by different users. The more agreed upon
translations it contains the more likely it is to be correctly understood.
(not just literally ;-)) (016)
I am not sure how all this can be transposed into your environment. I know
that there are groups currently trying to define an environment around RDF
that allows Multilanguage support. The system I work with was designed from
the beginning to incorporate the language support. I'll be happy to share
observations but I'm not sure if and when all of these can be part of the
currently available tools. (017)
What for:
Ontology based machine translation.
Explanatory systems describing terms though their underlying semantic
network.
Recognition of complex concepts in text (e.g. comparison of legal text
segments and graphical indication of their differences).
Comparison of statements to known knowledge and indication of conflicts.
. (018)
Tom (019)
-----Original Message-----
From: ontolog-forum-bounces@xxxxxxxxxxxxxxxx
[mailto:ontolog-forum-bounces@xxxxxxxxxxxxxxxx] On Behalf Of John F Sowa
Sent: Thursday, November 29, 2012 6:11 AM
To: ontolog-forum@xxxxxxxxxxxxxxxx
Subject: Re: [ontolog-forum] Concepts, ontologies, and natural languages (020)
The discussion about how NL terms are related to concepts, ontologies,
software, etc., is going around in circles without any criteria for
convergence or applicability. (021)
The word that is causing the most confusion is the word 'concept'.
It raises endless debates about how concepts are or are not related
to NL words and/or ontological categories and/or labels in some
computer software. (022)
In writing about conceptual graphs, I stated some criteria that
have successfully eliminated endless debates (at least for the
CG community): (023)
1. A concept is formally defined as a node in a conceptual graph (CG). (024)
2. The meaning of a concept is completely determined by the formal
operations that may be performed on the CG in which it occurs. (025)
End of debate. All further discussion is focused on the operations
for relating concept nodes to each other, to natural languages,
logics, ontologies, etc. Nothing else is relevant. (026)
For continuing debates in Ontolog Forum, I'll use similar criteria to
raise questions like "So what?", "Who cares?", and "What's the point?" (027)
John (028)
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J (029)
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J (030)
|