W3C home > Mailing lists > Public > public-sw-meaning@w3.org > October 2003

Re: A bit of running code for "Tim's View"

From: Sandro Hawke <sandro@w3.org>
Date: Wed, 08 Oct 2003 14:28:28 -0400
Message-Id: <200310081828.h98ISSW3024160@roke.hawke.org>
To: Jim Hendler <hendler@cs.umd.edu>
Cc: public-sw-meaning@w3.org


> if the bio:page states
>      bio:Dog a bio:mammal
               ^ assuming you mean ferrell:subClassOf   ?
> will you output
>   Donny a bio:mammal.

Yes.  The triples are all just throw together into one store.  As each
one gets thrown it, it gets examined to see if it suggests still more
should be thrown in.  (forming the "web closure", pruned (to be fairly
small in this case) because we're not following URIs in the subject
position.)  In cwm's case, given --think, once its done finding the
pruned web closure, then it computes the deductive closure (also
finite, even small), which it outputs.

There are some interesting possibilities for backward chaining, and
for reaching out onto web at the same time as doing inference, but cwm
doesn't try that right now.

The actual test I proposed for Tim ended with asking cwm which dogs it
knew about, so I didn't have to wade through all the other stuff cwm
might have gathered and inferred.

> If the bio: page contains
>    :Coconut a :Cat
> 
> will cwm claim
>    bio:Coconut a bio:Cat.
> ?

Yes.  The triples on the bio page are thrown in with the rest.  (I
suggested we may want to tag where they came from for
explanation/trust reasoning, some day soon.)

    -- sandro
Received on Wednesday, 8 October 2003 14:28:20 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:42:15 GMT