- From: Pat Hayes <phayes@ihmc.us>
- Date: Sat, 24 Nov 2018 15:08:51 -0600
- To: David Booth <david@dbooth.org>, W3C Semantic Web IG <semantic-web@w3.org>
- Cc: Nathan Rixham <nathan@webr3.org>
On 11/22/18 8:42 PM, David Booth wrote:> On 11/22/18 7:31 PM, Nathan Rixham wrote: >> . . . Graph = Set of Triple >> Triple = Node, Node, Node >> >> Why complicate it further? > > To *simplify* life for users. I think there is a *very* strong > case for supporting higher-level constructs like n-ary relations > and proper support for lists (unlike the hobbled RDF lists that > we currently have). These data structures are commonly used in > applications, and it just makes users' lives *more* complex if > they have to hand build these constructs instead of having them > directly supported in a higher-level language, ready to go. The > enormous popularity of property graphs is pretty strong evidence, > I think. I agree. But... > > However, I think the triple (or quad) model is a very good > foundation to build on. What I'd most like to see is a > higher-level RDF language that gets compiled into triples/quads, > just as python gets compiled into byte code, such that RDF users > never need to actually see or deal with the underlying triples. ...this analogy with high-level code being compiled into machine language is horribly misleading. There is nothing in the RDF spec to suggest that it is anything like a universal KR language in the same way that assembly code is a universal computer language. Higher-level constructions added to RDF will require semantic extensions to define their meanings, so even if you could compile them into RDF syntax, it would not be raw RDF. RDF simply is not expressive enough to handle what you want to say in it. It would be if it had negation and a way to indicate bnode identifier scopes explicitly (see my old 'blogic' slides for details: https://www.slideshare.net/PatHayes/blogic-iswc-2009-invited-talk slide 27 et seq) but this really WOULD be a major change to RDF. Pat Hayes > > David Booth > > > > -- ----------------------------------- call or text to 850 291 0667 www.ihmc.us/groups/phayes/ www.facebook.com/the.pat.hayes
Received on Saturday, 24 November 2018 21:09:19 UTC