SUO: Good for SUMO, Good for OpenCyc, good for all of us?
I agree with John that there seems to be no need to split
the IEEE mandate into more than one "SUO". From all that
has gone on before, I expect that for the foreseeable future
there will be a SUMO developed primarily by Teknowledge and
an OpenCyc developed primarily by Cycorp. These can continue
as ontologies used by their creator organizations for their
own purposes and also as candidate SUO's. John's motion
as I interpret it suggests a third candidate for IEEE-SUO,
which would be a merged ontology taken from SUMO, OpenCyc,
and with other content suggested by contributors. The
third ontology will have mappings to SUMO and OpenCyc and
will present an opportunity to test the utility of the IFF
theory for organizing multiple ontologies, and
an opportunity for those who would prefer not to
use either SUMO or OpenCyc to try to create something
better than either (whether better for some specific
purpose or better in being able to map more precisely
to more ontologies). Suggestions for content for the
merged ontology can be adopted or not into SUMO or
OpenCyc as their creators see fit. If the members of
Teknowledge and Cycorp find that additional effort to help
create a merged ontology is too distracting from their
immediate concerns, then the process will proceed more
slowly but can still proceed. If this group will undertake
to develop a merged ontology, I will certainly have
suggestions for additions of content not now in either
OpenCyc or SUMO.
Pat
============================================
John F. Sowa wrote:
> Adam,
>
> I don't understand why you insist on splitting
> the group. The content requires the methodology,
> and the methodology requires the content.
>
> Now that we have two separate content groups
> -- OpenCyc and SUMO -- it is even more imperative
> that we address the methodology of how these
> two ontologies can work together and with all the
> other ontological resources (WordNet, EDR, MeSH,
> UMLS, etc., etc., etc.) that any proposed
> standard will have to accommodate.
>
> Both SUMO and OpenCyc have developed separately
> for quite a long time, and they have made many
> decisions that have to be resolved if they are
> ever to be brought into a single framework.
> And the sooner we address those issues, the
> better it will be for everybody.
>
> If we can't get a consensus of the developers
> in this small committee, then there is no hope
> of convincing the rest of the world.
.
> John
--
=============================================
Patrick Cassidy
MICRA, Inc. || (908) 561-3416
735 Belvidere Ave. || (908) 668-5252 (if no answer)
Plainfield, NJ 07062-2054 || (908) 668-5904 (fax)
internet: cassidy@micra.com
=============================================