Re: Implementing statement grouping, contexts, quads and scopes

>On 2002-06-25 16:08, "ext Alberto Reggiori" <areggiori@webweaving.org>
>wrote:
>
>>  Patrick Stickler wrote:
>>
>>>  On 2002-06-24 18:00, "ext Jonathan Borden" <jonathan@openhealth.org> wrote:
>>>
>>>  Note that the approach I am proposing also would support arbitrary
>>>  "colored" triples, not just a binary unasserted/asserted distinction:
>>>
>>>  <rdf:Statement rdf:about="#foo"/>
>>>    <xxx:bar rdf:resource="#bas"/>
>>>  </rdf:Statement>
>>>
>>>  <rdf:Description rdf:about="&xxx;bar">
>>>     <rdf:type rdf:resource="#redPredicate"/>
>>>  </rdf:Description>
>>>
>>>  <rdf:Statement rdf:about="#foo"/>
>>>    <yyy:bar rdf:resource="#bas"/>
>>>  </rdf:Statement>
>>>
>>>  <rdf:Description rdf:about="&yyy;bar">
>>>     <rdf:type rdf:resource="#bluePredicate"/>
>>>  </rdf:Description>
>>>
>>>  <rdf:Statement rdf:about="#foo"/>
>>>    <zzz:bar rdf:resource="#bas"/>
>>>  </rdf:Statement>
>>>
>>>  <rdf:Description rdf:about="&zzz;bar">
>>>     <rdf:type rdf:resource="#greenPredicate"/>
>>>  </rdf:Description>
>>>
>>>  etc. where a given application (e.g. via a "color"
>>>  specific API) would interact with the knowledge
>>>  base, seeing as asserted both the fundamental
>>>  RDF asserted triples as well as the colored
>>>  triples of interest.
>>
>>  could you  try to make an example which uses the above syntax to encode a
>>  rule/formula? i.e. X=>Y
>
>I'm not sure I could, perhaps someone else might dare to craft such an
>example (if it can be done ;-)

Rules *cannot* be expressed in current RDF(S). That is, not without 
imposing some new semantic constraints that are not present in RDF(S) 
itself. That is a provable consequence of the RDF MT, so please 
everyone stop trying to do the impossible. N3 is not RDF(S), it is a 
semantic extension (in fact, an incompatible semantic extension - a 
different language -  since things like log:implies are not 
relations.)

>
>I don't see the definition of rules/formulas as the key purpose for
>this approach. But rather simply a way of expressing unasserted statements,
>which may or may not have additional qualifications per the fact that
>the construction is a reification (stating).
>
>>  can your syntax "scope" asserted and un-asserted as well then? i.e. doing
>>  things similar to the rdf:bagID effect :-)
>
>Well, that's more than I had anticipated having added to RDF parsing
>(since we want to be *very* conservative at this stage in the game)
>but that said, it would seem to me to be quite reasonable
>to have a similar interpretation of rdf:bagID on an rdf:Statement
>element per the proposed contracted form, to group the reified triples
>just as for rdf:Description.
>
>Of course, we could also simply say that RDF as it is now defined
>already provides for unasserted (dark) triples,

First, it does not, since the reified triples are not even present in 
the graph; second, your whole technique is irrelevant in any case 
since most OWL assertions cannot be encoded as a single RDF triple.

Pat


>and the nasty,
>obese syntax will just have to be tolerated until RDF 2.0 ;-)
>
>Cheers,
>
>Patrick
>
>--
>               
>Patrick Stickler              Phone: +358 50 483 9453
>Senior Research Scientist     Fax:   +358 7180 35409
>Nokia Research Center         Email: patrick.stickler@nokia.com


-- 
---------------------------------------------------------------------
IHMC					(850)434 8903   home
40 South Alcaniz St.			(850)202 4416   office
Pensacola,  FL 32501			(850)202 4440   fax
phayes@ai.uwf.edu 
http://www.coginst.uwf.edu/~phayes

Received on Tuesday, 25 June 2002 12:56:10 UTC