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

SUO: RE: RE: On the relevance of EXPRESS, EPISTLE, etc. to SUO -- RE: RE: RE: Some Procedural Suggestions




Dear Philip,

See comments below.

Regards  
      Matthew
============================================
Matthew West
Operations & Asset Management
Shell Services International
H3229, Shell Centre, London, SE1 7NA, UK.
Tel: +44 207 934 4490 Fax: 7929 
Mobile: +44 7796 336538
E-mail: Matthew.R.West@is.shell.com
http://www.shellservices.com/
============================================

> -----Original Message-----
> From: Philip Jackson [mailto:phil.jackson@computer.org]
> Sent: 09 February 2001 13:11
> To: West, Matthew MR SSI-GREA-UK; Standard-Upper-Ontology (E-mail)
> Cc: John F. Sowa
> Subject: RE: RE: On the relevance of EXPRESS, EPISTLE, etc. to SUO --
> RE: RE: RE: Some Procedural Suggestions
> 
> 
> Dear Matthew,
> 
> Please see responses below--
> 
> Regards,
> 
> Phil
> -------------------------------------------------------------------
> "Imagination is more important than knowledge. Knowledge is
> limited. Imagination encircles the world." - Einstein
> -------------------------------------------------------------------
> Standard Disclaimers. www.philjackson.prohosting.com
> 
> 
> > MW: If I compare EXIST and KIF I think they are functionally
> > fairly similar.
> > EXIST is designed with a processing model in mind, whereas 
> KIF is designed
> > for exchange. I also think that EXIST is easier to read 
> than KIF. This was
> > the main reason for rejecting KIF when we started to develop
> > EXIST. However,
> > I am aware that EXIST could be improved too from this point of view.
> 
> PJ: I think that "understandability" should be a criterion in 
> selecting a
> language, though it is subjective...People tend to like what they've
> developed or already learned...

MW: However, if we are to be successful inth elong run, we have to make what
we are doing accessible. Graham has talked much about using ACE, or a
restricted form of ACE as one notation for the SUO.
> 
> PJ: This reminds me of the issue that is often discussed 
> regarding the ease
> of understandability and expressiveness of predicate calculus 
> vs. natural
> language -- In particular, John Sowa has written about this issue in
> comparing predicate calculus vs. conceptual graphs... It 
> seems possible that
> the same kinds of arguments might translate to KIF (with the 
> same kinds of
> rebuttals)...(?) This may be a way to address the 
> understandability and
> expressiveness question...
> 
MW: Well there is another issue here, and that is that CG's have a
linguistic bias, which presupposes something of the nature of the ontology
you are going to create. Of course EXIST does the same thing with a
different bias.
> 
> > MW: So yes EXIST could make a good starting point for the 
> SUO. However, I
> > find here an assumption that KIF is the starting point, and I have
> > considered it unwise for a newcomer like me to challenge this.
> 
> PJ: Hmmm....I think it is generally wise at the outset of a 
> great endeavor
> to question all assumptions...

MW: I think I have done (more than) my share of questioning here.
> 
> PJ: This gets back to the issue, what are the requirements 
> for the ontology,
> and based on that, what are the requirements for the language 
> in which the
> ontology is expressed... Based on those requirements, which 
> languages best
> meet the requirements?
> 
> PJ: KIF is elegant, and a lot of thought and work has gone 
> into it, clearly.
> Yet it would not hurt to compare tools -- perhaps one or both 
> tools may
> benefit...
> 
MW: Well I'm sure EXIST would benefit from some critical review.
> 
> > > > > Also, I would like to make the suggestion that the SUO group
> > > > > should consider
> > > > > leveraging EPISTLE and other models already developed in
> > > EXPRESS as a
> > > > > starting point for developing SUO.
> > > >
> > > > MW: This is a better idea. The EPISTLE Core Model is
> > > designed to be a
> > > > foundation that can integrate a range of ontologies (or data
> > > > models). It is
> > > > not perfect, but it is the current result of some 15 years
> > > > concerted effort
> > > > by a range of people.
> > >
> > > PJ: I am glad to hear this. I hope that SUO'ers will give 
> very serious
> > > consideration to use of EPISTLE as a starting point.
> >
> > MW: I really think we would get there quicker if we did.
> 
> PJ: Well, this recommendation is now on the table, for the group to
> consider.
> 
> [...]
> > >
> > > > > EXPRESS, EPISTLE and the other models developed in
> > > EXPRESS, have the
> > > > > advantage of having been developed over several years by many
> > > > > people in an
> > > > > open, standards-based effort.
> > > >
> > > > MW: One of the key things we can bring is experience of
> > > > developing standards
> > > > in this area. The SUO does not have to suffer all the
> > > mistakes that we
> > > > managed to make.
> > >
> > >
> > > PJ: I think this is a very important point. Any insights 
> that you can
> > > provide should be very helpful to the group.
> >
> > MW: Most of the objections I have been raising to our 
> current course are
> > based on this experience. I know exactly where we are 
> currently headed -
> > I've been there.
> 
> PJ: Can you summarize from your perspectives, what mistakes 
> are currently
> being made by SUO (or towards which SUO is headed) and what you would
> recommend to address them?
> 
MW: The main lesson is that there are no short cuts, only diversions and
dead ends.
>