SUO: RE: KIF syntax and semantics and a basic ontology
Dear Chris (and Coworkers),
The syntax and semantics you have developed (not repeated below for brevity)
is an impressive piece of work, congratulations and thanks.
I have a number of questions and a few comments.
1. I feel rather like someone who has just been given the answer (which we
all know is 42) but not the question.
It would help greatly if you could document the requirements you are trying
to satisfy. This would mean we could check both that we could check whether
the requirements match our (individually and collectively) requirements, and
whether what you propose satisfies them.
2. You start by saying that instance-of is not required, and then in the set
theory bit go on to use it. Is this habit? Or are we to have two ways to
denote classification?
3. You have classes as non-extensional, i.e. two classes can have the same
membership but not be identical. I'm afraid this does not compute with me. I
was also not aware of a demand for this (whose requirement is this? what are
the benefits?)
However, I don't mind as long as I can have classes that are extensional -
which are presumably a subset of the ones that aren't. Could you explain how
I can get there?
If classes are non-extensional then I presume membership can also change
over time. There needs to be something that explains how you get to the
membership at some point in time.
If classes are non-extensional then I don't see what point there is in
having null. Since any class that had no members would have to be declared
as such, and not set equal to null, when would you use it?
4. There are a number of relations that appear to be undefined. These
include:
- documentation
- instance-of
5. Please can you give a bit more explanation about functional relations?
6. I don't understand what unary is really about. Do you mean singular, i.e.
there is only one member? Do you mean just one placed relations (classes?)
(though you say for n-tuples where n>2, rather than n>=2)? Or do you mean
something else? (please don't expect me to decifer the KIF).
7. You documentation for ternary says it is documentation for binary.
8. You have used the word sequence in a restrictive sense. There are other
sorts of sequence than number sequence. I suggest changing sequence to
number-sequence.
9. Signature looks much better than nth-domain, though I notice that with
the earlier changes you have made to KIF you can define even Nth-domain,
where it not being definable was my principal objection. Why have both?
10. I don't understand the last theorem about nth domain.
I haven't read all the stuff on arithmetic. I'll try to read that later.
Regards
Matthew
===============================================================
Matthew West http://www.matthew-west.org.uk/
Principal Consultant Shell Visiting Professor
Operations & Asset Management The Keyworth Institute
Shell Services International The University of Leeds
http://www.shellservices.com/ http://www.keyworth.leeds.ac.uk/
H3229, Shell Centre, London, SE1 7NA, UK.
Tel: +44 207 934 4490 Fax: 7929 Mobile: +44 7796 336538
===============================================================