- From: Peter F. Patel-Schneider <pfpschneider@gmail.com>
- Date: Sun, 7 Feb 2021 09:59:30 -0500
- To: public-rdf-star@w3.org
On 2/7/21 5:36 AM, Pierre-Antoine Champin wrote: > On 05/02/2021 17:17, thomas lörtsch wrote: >> It is easy to come up with scenarios where it is either important to record exactly which IRI was used to refer to :Berlin or where it is not important at all. > Absolutely >> The latter is most probably the normal case (and the proposed RDF* semantics doesn’t cover it). > > Can we agree that the proposed RDF* semantics covers it, although not in your favourite way? I'm not agreeing on this. There are https://w3c.github.io/rdf-star/tests/semantics/manifest.html#opaque-iri and similar test cases which, if approved, would require the RDF* semantics to be (semi-)opaque. But perhaps the question is whether it is easier to transform a transparent semantics to a (semi-)opaque semantics or vice versa. Let's consider a transparent semantics - standard RDF reification. I have shown that it is easy to transform standard RDF reification to opaque semantics of various forms by adding extra triples to the reification to encode the syntax of the triple that is being reified. This can be modified to be sensitive to the predicate of the triple that the embedded triple occurs in. It is also possible to add semantic rules to (semi-)opaque RDF* semantics that have triples in their interpretations. The rule if I(z) = I(z') then I(<< x y z>> :on w.) is true iff I(<< x y z'>> :on w.) is true adds a (little) bit of transparency. But these are only examples of the axiom that modifying the formal underpinnings of a logic can do just about anything. As far as I am concerned, none of the proposed RDF* semantics cover the transparent reading of embedded triples. peter
Received on Sunday, 7 February 2021 14:59:46 UTC