Dear Jim,
Of course, I am fully in favour of any possible, additional, serious work about the SUO topic because this, anyway, will increase our general knowledge about the ontological stuff. And I don't have, personally, any problem with DoD - I have just stopped working on an application of NKRL techniques on Afghanistan (declassified) news... My perplexities are linked mainly with the methodology chosen. As a Computer Science man, and not a philosopher, I am a little bit suspicious about any a priori, "in the abstract" comparison of methods. I prefer then to start bottom-up trying to find (implemented) ways of solve practical problems that can be, in case, generalized. If I remember well, DoD has already carried out in the past (2004) an a priori comparison of ontological Upper Level proposals (see the final report by Semy, Pulvermacher and Obrst): their conclusions were then in favour of DOLCE. Are they now unhappy with their 2004 findings?
Best regards,
G.P. Zarri
Schoening, James R Mr CIV USA AMC a écrit :
Gian Piero ZARRI,
Are you suggesting a given organization should not explore using an upper ontology? I agree there never will be a single correct upper ontology, but might there be a good-enough upper ontology that people or systems could agree to utilize?
Jim
-----Original Message-----
From: Gian Piero Zarri [mailto:zarri@noos.fr]
Sent: Thursday, March 20, 2008 9:20 AM
To: Schoening, James R Mr CIV USA AMC
Cc: standard-upper-ontology@LISTSERV.IEEE.ORG
Subject: Re: U.S. DoD WG expresses interest in upper ontologies
Dear Jim,
I think that my position on the subject is known: given the conceptual differences among the "upper levels" of models like DOLCE, CYC, Conceptual Graphs, SUMO, NKRL etc., the search for a Standard Upper Ontology seems to me like the search for some sort of ontological Holy Graal. I have devoted a whole Section to this topic in my forthcoming book on "Representation and Management of Narrative Information" (July 2008: some preliminary information about the book can be found on the Springer site: www.springer.com/computer/artificial/book/978-1-84800-077-3).
If the search for a SUO is, in my opinion, a false problem - let the market make the difference - a real ontological problem concerns, again according to me, the insufficiency of the present W3C tools (RDF(S), OWL) to represent and manage correctly many important "dynamic" situations like those described in "narrative" documents as the news stories, the medical records, many intelligence messages, the actuality photos etc. Note that a majority of the ontological gurus are perfectly aware of this but there exists, apparently, a sort of "psychological subjection" to the options defended by the W3C and by some EC mediocrities hidden behind the Luxembourg's hills: these suggest to search the solutions to this sort of problems within the present (W3C) paradigm and not outside it. But this is, of course, another matter...
Best regards,
Gian Piero ZARRI
Schoening, James R Mr CIV USA AMC a écrit :
All,
The U.S. DoD Ucore WG is exploring upper ontologies and is
interested in the pros/cons of those that exist. Their interest is
posted at http://data.army.mil/universal_core_wg.html and inserted
below.
SUO WG started an evaluation some years ago, which is still
posted at http://suo.ieee.org/SUO/Evaluations/, but it needs better
structure and more evaluation inputs.
If U.S. DoD were to select and start using an upper ontology,
this could create market momentum that could spill over into private
industry and lead to a global standard. Some might view this as good,
and others will be opposed, which is all fine. Those who would like in
influence any such selection should participate in this evaluation.
This evaluation should not just be for this potential user, but for any
and all potential users.
I should explain my role in this. I'm now working in the U.S.
Army's Net-Centric Data Strategy Center of Excellence. I was not the
person to suggest DoD Ucore WG make use of an upper ontology, but I
suggested they ask industry groups for evaluations of available upper
ontologies, and I offered to have my office host the public web site for
this announcement. I will not have much time to participate in this
evaluation, but I'm trying to encourage some young ontology developers
in my office to get involved. So, this is up to the members of this
group who see the opportunity in this statement of interest.
This is repeating myself, but should be restated: This group
does not currently have an official chair or other leaders. I've
suggested all along we wait until we have some real activities before
electing leaders, so if this creates greater activity, it may be time to
set up a leadership structure and elect leaders.
Jim Schoening
=========================================================
Below is text of statement posted at
http://data.army.mil/universal_core_wg.html
=========================================================
U.S. Department of Defense
Universal Core (UCore) Working Group
Statement of Interest in Upper Ontologies
The DoD Universal Core Working Group is exploring the use of an upper
ontology in conjunction with its Universal Core (UCore) data model. We
are interested in the advantages and disadvantages of each available
upper ontology, and we suspect other projects may also be interested.
Additional information can be found by searching on "Universal Core"
+DoD.
Groups and individuals with expertise in this field are invited to
conduct and share their analysis of available upper ontologies.
We are not requesting a recommendation, but rather an analysis of the
advantages and disadvantages of available upper ontologies.
We encourage interested groups and individuals to publicly post their
analysis and to provide a link to this page so all can share in the
results.
Links to Groups
Links to Individuals
Links to Papers and Studies
Links to Forums & Blogs
This page is provided by the U.S. Army CIO/G-6 Net-Centric Data Strategy
Branch in support of the DoD UCore Working Group.
Website Disclaimer | Privacy and Security | Accessibility Statement
|