Re: [External] : Re: handling ("asserted") s-p-o triples and ("stated" or "reified") id-s-p-o 4-tuples in RDF/SPARQL

Let me be clear about that: SPARQL treats rdfstar:states just like any other predicate, without any special assumption, since BGP matching is based on simple entailment, which is defined clearly in the baseline. Only under special entailment regimes, SPARQL may consider a special semantics for rdfstar:states, and this may be an extended RDF entailment or an extended RDFS entailment.
(This only if we really want a special rdfs:states predicate, about which I am more and more skeptical)
--e.

> On 15 Aug 2024, at 10:00, Peter F. Patel-Schneider <pfpschneider@gmail.com> wrote:
> 
> On 8/15/24 03:44, Thomas Lörtsch wrote:
> [...]
>> If an RDF-star entailment regime was enforced, in which 'rdfstar:states' entails the triple it states, then the stated triple term would be added to the graph as a proper/standard RDF triple, and consequently no additional BGPs would be needed that UNION a { :s :p ?o . } BGP with a  { ?id rdfstar:states <<( :s :p ?o )>> } BGP. As SPARQL is implementing RDF simple entailment, no such entailment of a triple from a stated term is available, and the above described UNION operation is needed instead.
> Entailments do not "add" anything to the graph.  The graph is the graph, after all, and is immutable.  Entailments instead show what follows from the triples in a graph.  Even in a system, an entailment regime can be enforced without adding anything to the data stored by the system.
> 
> peter
>

Received on Thursday, 15 August 2024 14:53:57 UTC