Re: Logic and Using The Semantic Web Toolbox

[Graham Klyne:]
> I haven't had the benefit of direct conversation with TimBL on this
topic,
> but based on his design issues "toolbox" paper I *think* he is claiming
> that negation can be defined using reification *and* introducing some new
> externally-defined concept;  i.e. that base RDF alone is not
> sufficient.  Which is what I understand you to be saying.

I agree: it does seem to me as if the key logical concepts intorduced by
TimBL are a "new externally defined concept", and as such would represent
the next step on from ontology; i.e. logic:-

     RDF => Ontology => Logic
        - http://www.w3.org/2000/Talks/1206-xml2k-tbl/slide10-0.html

My original question (some time ago), was how does logic *build* upon the
fundamental concepts of (firstly) RDF through reification, and ontology?
Onwards:-

[Pat Hayes:]
> reification provided some kind of general-purpose mechanism
> for semantic extensibility. If this is not the case, there
> seems to be little reason to have it in the language (?).

I still believe that essentailly there must be a "hack" using M&S and/or
ontologies to explicitly define what is meant by negation, and hence the
other forms of logical conept such as NAND etc.

Looking at http://www.w3.org/DesignIssues/Toolbox.html again...
Tim says that "not" is an XML element (containing, therefore, no semantics,
other than by human deduction?) and "truth" is taken as being an RDF
Property. He states that "Axiom of 'truth': strict alternative to not;
Semantics of 'truth': asserts boolean truth/falsity of document part". If
so, we are looking at a simple RDF Schema for boolean logic, and then using
boolean truth (1 or 0) to assert truth.

[Tim Berners-Lee (ibid.):]
> As RDF has little power at its basic level, anything new has to
> be introduced by reification - by describing it in RDF. Hence, to
> say "not(node, property, value)", you have to say, for example,
> "there is something which is an RDF property and has a subject
> of A and whose B property has vale C and is false".

Tim's idea to reify "not" using the truth property would be something
like:-

 <rdf:description about="#foo"
       xmlns:toolbox="http://www.w3.org/DesignIssues/Toolbox.html">
    <toolbox:truth rdf:value="0"/>
 </rdf:description>

Where the property of truth is defined in the SW Toolbox only...
The odd thing about this that follows is that Tim doesn't bother reifying
anything (any "not" statement) anymore, he just uses the <not> element.
Maybe he did that for brevity, and maybe he did that because he thought it
was a fundamental SW concept missing from the earlier levels of SW
architecture. It is that which I am currently uncertain about.

Also keep in mind that the <not> assertion is a very powerful and
fundamental piece of logic indeed...

Happy holidays everyone!
Kindest Regards,
Sean B. Palmer
http://uwimp.com/
http://www.w3.org/WAI/ [ERT/GL/PF]
"Perhaps, but let's not get bogged down in semantics."
   - Homer J. Simpson, BABF07.

Received on Saturday, 23 December 2000 11:15:37 UTC