[om-list] Regular Expressions

[Thomas L. Packer at home]

Hello Rajeev

    I'm not familiar enough with most standard things like regexp.  I'm only
slightly familiar with frame-based representations, and when I studied them,
I remember I felt like they were arbitrary and incomplete.

    I'll have to research ASP, because I'm not sure what it stands fro.  I
don't use ASP.NET.  I've stolen the same acronym to mean "Abstract Sequence
Patterns", which is what I've been calling an expression written in my
regular expression language.  But I may have to change my names if they are
ambiguous.

    Oh, you must be reading the older OM-list emails.  That's where you came
across my use of "ASP", right?  Good, good.

    As far as representing language, I think a recursively-defined abstract
sequence representation will work quite well, but I'll have to prove it.
But that's just for language-expression representation.

    I've been thinking about a knowledge modelling language for much longer
than I've been thinking about regular expressions, and I'm sure that frames
are not good enough, nor is anything else I've seen.  What I have in mind
for that, which is a long way from being developed, and is not really
related to my ASPs (yet), is what I call a meta-relational complex
hierarchy.  It uses "nodes" to represent arbitrary ideas, it uses
"internodes" to represent relations between nodes, and it uses
"meta-internodes" to represent what I all meta-relations, or relations
between relations.  I started to develop other elements that seemed
necessary at the time, but it's been a while since I thought about this in
any detail.  But I'm pretty sure once I re-think things, I'll be convinced
again that this three-part knowledge representation is necessary (and
perhaps sufficient) to representing any type of knowledge -- the three
parts, plus a few other principles that I was writing about in an essay
called the MetaMathesis Essay, which outlines what I think is necessary and
sufficient for KB software design.

    The other principles include the proper use of a number of unavoidably
interdependent dichotomies, such as quantity-quality, induction-deduction,
static-dynamic, etc.

    To do knowledge modelling right, I think we will need to recognise and
make use of quite a few principles, many of which are either totally
ignored, or are included accidentally and without proper discipline, by all
the KB forms I've heard of.

    ciao for now
tomp

~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Víðar sum quem nihil obstat.
www.Ontolog.Com
~~~~~~~~~~~~~~~~~~~~~~~~~~~~


----- Original Message -----
From: "MrM0j0r15n Rajeev the kanehbosm" <electrickanehbosm at hotmail.com>
To: <om-list at onemodel.org>
Sent: Saturday, 31 August, 2002 01:07
Subject: [om-list] Regular Expressions


>
> Hello tomp,
>
> the ASP idea that you had sounds suspiciously similiar to regexp ;)
>
> Anyway, is it actually possible to represent natural languages in terms of
> say ASP ? the ALICE program i referred to earlier
> www.alicebot.org, uses a similiar recursive reduction/expansion in its
> operations (and it won an annual prize, based on a Turing Test).
>
> What are the advantages of using ASP like language ? what about its
> advantages over something like a frame-based representation ? or is it
> equivalent ?
>
> insanekane
>
>
> _________________________________________________________________
> Chat with friends online, try MSN Messenger: http://messenger.msn.com
>
>
> _______________________________________________
> om-list mailing list
> om-list at onemodel.org
> http://www.pairlist.net/mailman/listinfo/om-list