Missing rule in closure step 5. To cover targets E that mention triple terms, step 5 must also apply Grdfs14. Without it, the procedure may fail to derive <<(s p o)>> rdf:type rdfs:Proposition, breaking the completeness claim for such cases.

Maybe I don't follow. The text clearly says that up to step 5 the rules are sound but not complete. This is true. At the end of the section we introduce the RDFS-T entailment pattern with rule Grdfs14 to recover completeness.

I was confused by that comment because point 5 appears twice in the section. Enrico is right for the first point 5, it is clearly stated afterwards that the procedure is not complete. For the second point 5 there is no claim for completeness either. But in that list, I think, we should take graph terms into account and we need to incorporate rule Grdfs14 in the list. Alternatively, we can make a third list with 5 points (even more confusing though) which takes graph terms into account.

Note that we also need to change
For every IRI or literal aaa used in E, add aaa rdf:type rdfs:Resource to S.
to
For every IRI, literal or triple term aaa used in E, add aaa rdf:type rdfs:Resource to S.

But now that I am starting with this kind of comments, I think we should systematically go through the section and adjust it once we agree that the semantics is finite.

Incorrect wording “symmetric RDF triples.” Triple terms are not “symmetric.” In legal RDF they occur only in object position; in the generalized setting you allow them also in subject position. Please rephrase accordingly (e.g., “generalized RDF where triple terms may occur in subject position”).

I believe that the current text is correct, given the definition of symmetric RDF triples. Am I missing something?

Here the confusion might stem from the fact that symmetric RDF is already a subset of generalized RDF. The statement is not wrong, but maybe confusing in the sense that there is not generalized RDF without symmetric RDF triples.

Error in the RDFS derivation (second example).> The rdfs7 step is wrong:> Current: ex:d _:b ex:c . by rdfs7
Correct: ex:d _:b ex:e . by rdfs7
(Then ex:d rdf:type ex:c . by rdfs2 follows.)

Your are right, this should be fixed.

Yes, that is simply a typo.

check correctness of Appendix A #139

Description

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions