- From: Jeremy Carroll <jjc@hplb.hpl.hp.com>
- Date: Wed, 6 Mar 2002 09:47:59 -0000
- To: <www-webont-wg@w3.org>
A small clarification, and then a more detailed reply. Clarification: ============== Jeremy: > > What I have seen so far is the desire to state logical entailments. > > > > This is quite addressable outside RDF rather than inside the graph. > > The N3 version could be modelled in RDF as a load of triples, with N3 > > contexts roughly corresponding to bags of reified triples in RDF. Jonathan: > Yuck! Oh yes I share this sentiment. What I was trying to say, but didn't, was that the bags of reifications is a possible (but unattractive) approach within RDF. But outside RDF it is also quite possible to state entailments. More detail =========== For example, we can have separate files, some of consequences, some of conclustions, and then say that files A, B, and C entail files D, E and F. This can be made formal quite easily. My belief is that that is sufficient for the only use cases I am aware of that is in charter for this group. Those use cases are to have clear discussions within the group and for the stating of test cases for implementations. Entailment & Logic ================== > Jeremy Carroll wrote: > > What are the use cases for this ability? > > Basically anywhere you see (in N3) { ... } a context is defined > in which the > contents are not necessarily asserted. > > Perhaps the simplest use of this is an IF THEN statement or an OR > statement > > IF {sky color blue} THEN {trees color green} [..snip..] > A simple way to write down a simple formula is the use case. My problem with this is whether the examples are in scope. I am in this group to help provide a language to specify ontologies for data and information on the web. N3 and the examples appear to be trying the rather more ambitious task of having a web language for writing complex logical expressions. A worthwhile task, but not my understanding of our task. I believe that we will need to talk about logical entailment. (In order for us to discuss and specify the meaning of OWL). I do not believe that we will need logical entailment in our language. Jeremy
Received on Wednesday, 6 March 2002 04:48:03 UTC