SUO: Re: Formal SUMO Draft -- *Date 01 Feb 2002
I support the comments by Jon and Pierluigi. In fact, the current
proposed standards for KIF and CGs have been combined to downgrade
syntax and highlight semantics.
The NWI (New Work Item), which will be presented to the SG32
committee in May, is a single standard for Logical Foundations (LF).
KIF and CGs are two different syntaxes, whose rules will be included
as external representations of the LF. However, there is nothing
magical or required about either KIF or CGs, and any other syntax
that can be formally mapped to the same LF would be equally suitable.
Jon Awbrey wrote:
>> In my opinion it is a fundamental mistake to specify
>> a particular logical language, for example, KIF or
>> any other, as a part of the compliance conditions.
>> This would be as bad a practice as stipulating
>> that a compliant ontology has to be written in
>> English as opposed to French or German, or to
>> use an even more notorious analogy, that
>> a program has to be written in ADA.
>>
>> Requirements should be specified at a higher level
>> of abstraction and generality than any particular
>> ontology oriented logical formalism.
That is exactly what we are proposing for the LF. Any notation
that has a formally defined mapping to LF is equally suitable
for representing ontologies or anything else that requires
a formal specification.
Pierluigi Miraglia wrote:
> Adding my $0.02, I agree with J.A.: I must say that I still have
> misgivings about the very idea of 'standard' ontologies all along, but
> aside from that it seems that the point of an ontology is primarily to
> serve as an abstraction layer for _content_.
>
> Perhaps a better compliance condition could be formulated in terms of
> traslatability or 'representability'. An ontology is compliant iff (i) it
> is formulated in _a_ language with features X, Y, Z and (ii) it is
> accompanied by a translation manual into the 'standard framework'.
I share those misgivings about 'standard' ontologies, but the issue
of standard ontologies is completely separate from the issue of
standard notations for ontologies. With the LF, the fundamental
logical features are specified in an abstract form that can be
expressed in two starting notations (KIF and CGs), but with no limit
on the number of additional notations (graphic, linear, or whatever)
that may be added.
I also agree with Jon's final comment:
JA>> Among the more deleterious side-effects of
>> thinking in only one language is a constant
>> tendency to confuse that language with reality.
John Sowa