issues related to RDF semantics

 Issues to discuss at the Semantics Task Force document on 7 March 2025.

I suggest that the meeting discuss the following issues, in this order:

vocabulary to refer to the individual nodes in a triple term #130 
https://github.com/w3c/rdf-star-wg/issues/130
   This could have impact on semantics if <<(:a :b :c)>> rdf:subject :a. is an 
axiomatic triple.
   I think this needs to be discussed in the meeting, but with a strict time 
limit of at most 20 minutes so that the other issues can be discussed.

A formal background to unify triples and triple terms #87 
https://github.com/w3c/rdf-semantics/issues/87
   I think that this proposal has flaws and should not be added to Semantics 
at this late stage.
   I think this needs to be discussed in the meeting, but with a strict time 
limit of at most 20 minutes so that the other issues can be discussed.

Un-star operation to support RDF Dataset Canonicalization? #114 
https://github.com/w3c/rdf-star-wg/issues/114
   I believe that this issue should not affect Semantics, either directly or 
indirectly through any changes that it would make in Concepts.  The meeting 
could vote to make this a requirement of unstar.

Define an interpretation of Triple Terms #49 
https://github.com/w3c/rdf-semantics/issues/49
   I believe that this issue is already handled by the semantics in Semantics. 
  Changes are, of course, possible.  The meeting could vote to close the issue 
based on this.

what properties can or should link to triple terms? #127 
https://github.com/w3c/rdf-star-wg/issues/127
   I believe that this issue has been resolved by the wording about 
rdf:reifies in Concepts and there are currently no implications on Semantics. 
The meeting could vote to recommend that the issue be closed based on this.

map the annotation syntax to rdfs:states #128 
https://github.com/w3c/rdf-star-wg/issues/128
   This is an alternative way to do annotated triples.  It's currently not in 
the documents.  I propose that the meeting vote to close the issue with no 
change to documents.

the change to replace "must" with "should" for ill-typed term values should be 
reverted pending further discussion #147 
https://github.com/w3c/rdf-concepts/issues/147
   The original issue is https://github.com/w3c/rdf-concepts/issues/60.
   This was discussed in 
https://www.w3.org/2024/10/31-rdf-star-minutes.html#783b.  There was a straw 
poll with mostly positive support and no -1 votes, but no proposal.
   This was again discussed in 
https://www.w3.org/2024/12/05-rdf-star-minutes.html#9859.  The vote there was 
to defer until some data came back from Pierre-Antoine.
   Pierre-Antoine did come back with information that QLever, Virtuoso, and 
Allegrograph do raise an error.  So does Blazegraph.  The test was not 
conclusive so some systems that were listed as not producing an error might in 
other circumstances.
   This was discussed in https://www.w3.org/2024/12/19-rdf-star-minutes.html#0af2
   It looks as if this does require some attention, but I don't see any effect 
on Semantics, except that the bit that talks about signalling an error for 
unsatisfiable graphs should be removed.
   I created https://github.com/w3c/rdf-concepts/issues/164 for the larger 
issue of unsatisfiable graphs.
   I believe that there are no effects to the semantics arising from this 
issue, as the RDF semantics is just about satisfaction and entailment and 
should not be concerned with other behaviour.

Explain how classic RDF reification relates to triple terms and rdf:reifies 
#61 https://github.com/w3c/rdf-semantics/issues/61
   My view is that this is not something that should go in Semantics.

completeness of entailment #76 https://github.com/w3c/rdf-semantics/issues/76
   I created a PR (#101) that qualifies everything.  I think we can merge the 
PR.  Hopefully we can iron out enough of the above issues that at least the 
interpolation lemma can be proven.

identity and equality of datatype values #92 
https://github.com/w3c/rdf-concepts/issues/92
   I think the task force should state that it is important to make clear what 
RDF equality means between literal values, i.e., that the denotations of +0 
and -0 in any RDF datatype that includes IEEE floating point are not equal as 
far as RDF is concerned even though they compare equal in IEEE floating point.



peter

Received on Thursday, 6 March 2025 18:47:20 UTC