Re: Decision from the Semantics TF: liberal baseline

I agree with everything Pierre-Antoine is saying here.
—e.

On 9 Jan 2025, at 11:57, Pierre-Antoine Champin <pierre-antoine@w3.org> wrote:



On 09/01/2025 10:47, Franconi Enrico wrote:
These are the more precise definitions separating the RDF and RDFS semantics.
They are already in the document
https://github.com/w3c/rdf-star-wg/wiki/RDF-star-%22liberal-baseline%22

Comments welcome.
RDF SEMANTICS

RDF semantics restricts the function [I+A](.) with the following additional metamodelling condition:

  *   <r, [I+A](rdf:Property)> ∈ IEXT([I+A](rdf:type))
          if ∃ x,y,z . RE(x,r,z)=y

I'm pretty sure this additional condition is not required, as it already follows from what we have.

if ∃ x,y,z . RE(x,r,z)=y,
then from the domain of RE, it follows that r ∈ IP,
then from the 1st semantic condition<https://www.w3.org/TR/rdf12-semantics/#rdfsemcond1> of RDF semantic, <x, [I+A](rdf:Property)> ∈ IEXT([I+A](rdf:type))

Additional metamodelling sound and complete RDF entailment patterns (informative):

if the triple structure appears in S    then S RDF(S) entails
reif0   sss aaa ooo     aaa rdf:type rdf:Property .
The entailement pattern above is indeed required, but I would propose it as a replacement of rdfD2<https://www.w3.org/TR/rdf12-semantics/#dfn-rdfd2> rather than a new separate pattern.
RDFS SEMANTICS

RDFS semantics restricts the function [I+A](.) with the following additional metamodelling conditions:

  *   ⏩ <r, [I+A](rdfs:Resource)> ∈ IEXT([I+A](rdf:type))
          if ∃ x,y,z . RE(x,z,y)=r or
          if ∃ x,y,z . RE(x,z,r)=y or
          if ∃ x,y,z . RE(r,z,x)=y ⏪️

Similarly, I don't think the above is necessary, which I briefly indicated already<https://lists.w3.org/Archives/Public/public-rdf-star-wg/2025Jan/0053.html> (but granted, this is a challenging thread to follow...). So, in more details:

*  if ∃ x,y,z . RE(x,z,y)=r , then per the definition of RE, r ∈ IR,
    RDFS semantic conditions<https://www.w3.org/TR/rdf12-semantics/#rdfssemcond1> tell us that
        - ICEXT(y) is defined to be { x : < x,y > is in IEXT(I(rdf:type)) }
        - ICEXT(I(rdfs:Resource)) = IR
    which implies that < r, I(rdfs:Resource) > ∈ IEXT(I(rdf:type))
* if ∃ x,y,z . RE(x,z,r)=y, then per the domain of RE, r IR,
   and the same reasoning as above applies
* if ∃ x,y,z . RE(r,z,x)=y, then per the domain of RE, r IR,
   and the same reasoning as above applies

the condition below, on the other hand, is required (with the caveat of deciding whether we want the last line or one -- rdfs:Proposition begin the "super range" of rdf:reifies).

  *   ⏩ <r, [I+A](rdfs:Proposition)> ∈ IEXT([I+A](rdf:type))
          if ∃ x,y,z . RE(x,z,y)=r or
          if ∃ x,y .   RE(x,[I+A](rdf:reifies),r)=y ⏪️

Additional metamodelling sound and complete RDFS entailment patterns (informative):

Similiar remark as about reif0 above (oh, and we should avoid giving the same name to two patterns!):
reif0 below is a generalisation of rdfs4a<https://www.w3.org/TR/rdf12-semantics/#dfn-rdfs4a> and rdfs4b<https://www.w3.org/TR/rdf12-semantics/#dfn-rdfs4b> so we should rather extend those than introduce a new one.

if the triple structure appears in S    then S RDF(S) entails
reif0   sss aaa ooo     sss rdf:type rdfs:Resource .
ooo rdf:type rdfs:Resource .

if S contains   then S RDFS entails
reif1   sss aaa <<(xxx yyy zzz)>>       <<(xxx yyy zzz)>> rdf:type rdfs:Proposition .
reif2   <<(xxx yyy zzz)>> aaa ooo       <<(xxx yyy zzz)>> rdf:type rdfs:Proposition .
reif3   sss rdf:reifies ooo     ooo rdf:type rdfs:Proposition .

Anyway, it is not so important to find the most compact set of patterns, since these patterns are (a) only informative, and (b) not the basis for any actual implementation.

I agree, even though I'm not sure about (b)...

  best