- From: Holger Knublauch <holger@topquadrant.com>
- Date: Fri, 23 Nov 2018 10:18:57 +1000
- To: semantic-web@w3.org
On 23/11/2018 10:08 AM, William Waites wrote: >> The prevailing thought at the time was that there was value in >> being able to make such "existence" assertions, so that is what >> we got in the RDF semantics. But after 20+ years of use, I think >> it has become clear that this subtle distinction is not actually >> *needed* in practice, as Skolem IRIs clearly demonstrate. > FWIW, I have a use-case in synthetic biology that does actually > use this distinction. If I were to use Skolem IRIs instead, I would > have to look into the IRI itself to find out if it was a Skolem > constant or a normal one or else invent a vocabulary for trying > to talk about statements (oops) and say that they’re meant to > express existence. That’s all less convenient and less clear than > checking if the assertion involves existential variables and > figuring out in an application-specific way what resources exist > that could be used to satisfy it. > > This is not to say that I couldn’t solve the problem in some other > very different way, but that there are use-cases where understanding > blank nodes as meaning existential quantification is actually used > in practice. Have you tried to express your use case in SPARQL? While I haven't been around for the original RDF semantics discussions, it feels a bit like they were attempting to mix data representation with a (simple) query and matching language. In retrospect, and with SPARQL now around, this was probably an unfortunate decision that caused a lot of unnecessary complexity. Holger
Received on Friday, 23 November 2018 00:19:25 UTC