[Aleader-dev] Re: more about KM

[William L. Jarrold]

On Tue, 19 Aug 2003, Joshua N Pritikin wrote:

> I didn't realize what you meant when you said about KM that:
> "It is also frame based.  This can cause problems."
>
> Ugh.
>
> The representation stuff and unification are cool.  BUT, trying
> to do any non-trival inferencing is a chore.

Yes.  This is par for the coarse for any KR system these days.  Maybe in
10 years there will be lotsa patterns and modules that we can use.
Porter et al have a paper called Knowledge Patterns.

>
> I mean, is it really true that I can only make prepositions
> like (:triple f s v)?  That's nuts.  I'd be forced to move
> half of the stuff into lisp and call back & forth.  Maybe
> that's what I'll end up with.

I'm not sure I understand what you are saying here.  There are
several possible interps.  One is that you have many non-binary
predicates in your model.  Is that what it is?

It is very likely that you are on the "same page" regarding this, but
just to be sure....The :triple thing is only necessary for certain stuff.
Many rules can be written without it....but as I said, you already know
that, right?

Anyway, this reaction of yours argues in favor of CycL.  I am much more
familiar with Cyc than KM.

Bill

>
> Do you have any experience with PowerLoom?  I'm going to play
> with it for a few days and see how it feels.

No experience.  Chances are pretty good that it sucks compared to CycL.
What is its licensing?

Bill

>
> --
> A new cognitive theory of emotion, http://savannah.nongnu.org/projects/aleader
>