Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

Re: SUO: RE: IFF Examples (was Question)




Matthew,

Thanks very much for your offer.

I am interested, and I'll be in touch with you about this
(off-list) soon.

Jim

On Wed, 10 Apr 2002, West, Matthew R SITI-ITPSIE wrote:

> 
> Dear Jim,
> 
> Let me make an additional offer.
> 
> I am one of the developers of ISO15926, which is a data model and
> reference data library that effectively makes an ontology (without
> many axioms) that is 4D in nature.
> 
> This could give a third point on your graph, from a different
> paradigm, purpose and language.
> 
> I would be happy to work with you to see how this could be brought
> into the IFF framework.
> 
> 
> Matthew West
> Principal Consultant
> Shell Information Technology International Limited
> Shell Centre, London SE1 7NA, United Kingdom
> 
> Tel: +44 20 7934 4490 Other Tel: +44 7796 336538
> Email: matthew.r.west@is.shell.com
> Internet: http://www.shell.com
> 
> 
> > -----Original Message-----
> > From: Jim Farrugia [mailto:jim@spatial.maine.edu]
> > Sent: 09 April 2002 19:31
> > To: SUO
> > Cc: Robert Kent; Leo Obrst; johd@cyc.com
> > Subject: SUO: IFF Examples (was Question)
> > 
> > 
> > 
> > As part of the IFF group, I am working on the following and 
> > will share 
> > progress and questions with the SUO group as the work takes shape. 
> > 
> > The goal of what follows is to provide suitable material for creating
> > examples for the IFF. 
> > 
> > 
> > 1) Extract a "suitable part" of the SUMO.
> > 
> > 2) Extract a "closely corresponding" part of OpenCyc vocabulary and 
> > assertions.
> > 
> > 3) Extract a "domain-specific" part of OpenCyc (probably something
> > related to NaiveSpatialMt or NaturalGeographyMt).
> > 
> > 4) Create a plausible "SUMO-like-but-lower-level" counterpart to
> > #3 above.
> > 
> > 5) Use #1 - #4 to create examples that show how the IFF 
> > accomplishes its 
> > goals.
> > 
> > Easier said than done, I'm sure. And we'll likely need to 
> > "add" certain
> > pieces that aren't extractable from the available mix.  But, ...
> > 
> > I have received some kind help from John D. at Cyc, and I'm 
> > now working
> > with the OpenCyc browser interface and the API to get what I need.
> > 
> > I expect it will take me quite a while.  
> > 
> > As I learn about how to do things in OpenCyc I will share 
> > this knowledge 
> > with the SUO group. And as I have questions about what I'm 
> > doing, I may 
> > lean on the group for some help.
> > 
> > Extracting a "suitable" piece of the SUMO will also not be trivial.
> > 
> > Please note two things:
> > 
> > (1) our goal is NOT to provide a _software implementation_ of the IFF;
> > 
> > (2) the extracted/created examples are not supposed to be yet another
> >     candidate ontology for the SUO group. Rather, they should 
> > simply be
> >     _plausible_ ontologies of the kind the IFF is designed to 
> > work with.
> > 
> > The goal is to create clean examples (for the IFF), with 
> > sufficient detail 
> > to be convincing, and of sufficient simplicity so as not to 
> > be overwhelming.
> > 
> > Given the availability of SUMO and OpenCyc, the above approach seems
> > (at firts glance) like a plausible way to approach things.  I 
> > hope it also 
> > proves generally useful to the SUO group.
> > 
> > Any and all comments welcome.
> > 
> > Jim F.
> > 
> > 
> > On Tue, 9 Apr 2002, John F. Sowa wrote:
> > 
> > > 
> > > Adam,
> > > 
> > > That's a good start:
> > > 
> > >  > SUMO is divided into 11 modules and the entire 15-month 
> > history of
> > >  > the development of SUMO continues to be available at
> > >  > 
> > <http://ontology.teknowledge.com/cgi-bin/cvsweb.cgi/SUO/Merge.txt>.
> > >  > Attribution of original merged sources is available in 
> > the version
> > >  > comments, in-line with the source code, and separately 
> > formatted in
> > >  > a document at
> > >  > <http://ontology.teknowledge.com/rsigma/annotatedHistory.html>
> > > 
> > > There are many next steps, but critical among them are
> > > 
> > >   1. Define a standard for the metalevel information (for which the
> > >      introspection information of Java beans would be useful as a
> > >      beginning).
> > > 
> > >   2. Analyze OpenCyc to see how much correspondence there is between
> > >      the SUMO modules and the OpenCyc modules.
> > > 
> > >   3. If necessary break down both the SUMO modules and the OpenCyc
> > >      modules into smaller ones in order to find matching 
> > and nonmatching
> > >      parts.
> > > 
> > >   4. Translate the CycL modules into KIF to make the comparison
> > >      easier (or better yet, get the Cyclers themselves to adopt
> > >      the Common Logic standard so that the translations could be
> > >      automated).
> > > 
> > > There's a lot more to do, but this enough homework for today.
> > > 
> > > John
> > > 
> > > 
> > 
> > 
> > 
> > 
> > 
> > 
> > 
> > 
> > 
> > 
> > 
>