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

Re: SWSL declarative semantics

From: Jim Hendler <hendler@cs.umd.edu>
Date: Fri, 21 Nov 2003 09:19:41 -0500
Message-Id: <p05200f15bbe3ce1b0011@[]>
To: "Peter F. Patel-Schneider" <pfps@research.bell-labs.com>, kifer@cs.sunysb.edu
Cc: bparsia@isr.umd.edu, public-sw-meaning@w3.org

At 8:49 AM -0500 11/21/03, Peter F. Patel-Schneider wrote:

>Hmm.  It would be interesting to see a pictorial formalism where this
>really works.  Pictures are worth a thousand words, of course, and thus it
>is usually the case that it takes more than a thousand words to really
>tease out what is going on in a pictorial formalism, which does not
>generally lead to nice logical formalisms.  Formalisms that start out
>pictorially usually end up abandoning the pictures except for
>non-authoritative display purposes.  (Semantic Networks, Frames, and
>Brachman's original structural inheritance diagrams have all gone this
>route.)  I remember arguments about what was really going on in KLONE-talk,
>a graphical interface to KL-ONE that was abandonded, largely because its
>manipulations of KL-ONE graphs did not correspond to reasonable
>modifications of the underlying formalism.

For what it is worth - most of the pictorial frameworks also get 
abandoned for two other reasons -- they don't scale well, and they 
cannot usually be emailed or otherwise generated quickly and easily 
and in a machine independent way that can be read using a textual 
interface.   Another reason is that as things get more expressive 
(and all these kind of langauges did) the pictorial framework suffers 
-- consider how to draw the difference between "allValuesFrom" and 
"someValuesFrom" in an intuitive way.   UML is still often rendered 
pictorially -- but users who use its advanced features complain about 
   The decision to move away from the pictorial examples in KL-ONE is 
a good example of where the neats and the scruffies both agreed it 
wasn't useable, but for somewhat different reasons.

  All that said, it doesn't actually have  much to do with the "social 
meaning" debate, so I'll shut up...

Professor James Hendler			  http://www.cs.umd.edu/users/hendler
Director, Semantic Web and Agent Technologies	  301-405-2696
Maryland Information and Network Dynamics Lab.	  301-405-6707 (Fax)
Univ of Maryland, College Park, MD 20742	  240-277-3388 (Cell)
Received on Friday, 21 November 2003 09:22:10 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 19:56:01 UTC