W3C home > Mailing lists > Public > www-rdf-interest@w3.org > March 2001

Re: Open Worlds, Distribution, Delegation, Federation, Logic, NI

From: Sandro Hawke <sandro@w3.org>
Date: Fri, 02 Mar 2001 13:31:18 -0500
Message-Id: <200103021947.f22JlF714676@daniel.hawke.org>
To: jos.deroo.jd@belgium.agfa.com
Cc: www-rdf-logic@w3.org, www-rdf-interest@w3.org

> I'm really with you on those topics, although I haven't
> had enough chance to talk with you about those topics on the
> technical pleanary meeting (but I have still your face
> in front of my mind ...).

Ah!  I never made the name/face connection for you until just now,
looking at your web page.   I thought some more about what you asked
after my little talk, but now I can't quite remember it.  (I've
learned a lot more about logic this week; some of it's pretty
tangled.)

> With respect to negation, I don't know if we are avoiding naf
> (negation as failure) when we are using
>   {:subject :verb :object} log:implies <data:,>
> which is a specific form of a nested implication (if I'm correct)
> The <data:,> (or the "" if you want) is not the same as
> log:Falsehood but just stands for a no-proof-found instance
> which could be the result of any (delegated) attempt to find
> any evidence (when using a universally quantified verb for the
> conclusion). So I'm not sure wether that's naf or not, but it
> certainly _opens_ the way engines can interact ...

This is the first I've heard of this convention.  Did you come up with
it for Euler?   

My approach (not yet implemented) is much more explicit: conclude
(instead of a literal) something like "this engine could not satisfy
this proof", which I can't think of how to express in n3 right now.
"This engine" is an identifier for the running process which tried to
find the proof, and "this proof" is an identifier for the proof
requested.  or something like that.

I think that this is all fine (and necessary), as long as we're
careful not to conclude Falsehood when it is not warranted.  In
particular, it may often be appropriate to infer something being false
if a particular engine, operating on a particular data set, was not
able to find a proof.  

But I think the NAF concept used more freely will lead people to
seriously overconstrain their system, stopping it from evolving.

While I'm following up my earlier message, I want to mention that
Harold Boley said my proposed system for communication with
meta-queries sounded like a "blackboard architecture" which indeed it
does.

     -- sandro
Received on Friday, 2 March 2001 13:31:39 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:51:48 GMT