Re: Annotations with RDF statements

Hi Antoine,


> On 2 May 2016, at 20:06, Antoine Isaac <aisaac@few.vu.nl> wrote:
> 
> Dear Web Annotation group,
> 
> (sorry if this comment has been made already, but I have no time to check...)
> 
> At Europeana we have a problem of having a annotation that is an RDF triple.
> Previously the Open Annotation model had a whole section about solving this issue by having RDF graphs as bodies:
> http://www.openannotation.org/spec/core/publishing.html#Graphs
> Actually *precisely* the issue we have now...
> 
> The last WA draft has nothing for it. It says in the change log:
> "Deprecate embedded graphs as an explicit part of the model, instead just include or reference a serialized graph."
> 
> I'm a bit disappointed by the fact there's no strong guidance anymore, as you can guess ;-)
> But I can understand the WG perspective: embedding graphs is a tricky matter.
> What I'm puzzled by is the text of the spec doesn't even mention something about "include or reference a serialized graph" anywhere, it seems. So even the weak guidance is not part of the spec anymore? Or am I missing anything?
> 

I do not think you are missing anything; there is no such example in the text afaik. The question is whether there should be or whether, instead, we should think about having a primer showing some of these examples and usages.

In many respect, by now, this issue became much less of an issue. Since 2014, ie, since the publication of RDF1.1, we do have the standard notion of RDF Datasets[1], or 'named graphs' in the usual RDF geek parlance. In other words what is described in the OA document, and which was slightly tricky back then, should not be considered as tricky any more: giving a URI to a triple(s) is properly defined in a standard way. In other words, I'm not sure there is a reason for calling out this particular usage pattern from among the many many different usages of URI-s for bodies and targets… except in a primer!

All that being said: a primer may be out of reach with the current timing and manpower. Adding such an example, as part of an informative appendix of the vocabulary document would be o.k. with me if people want that. But it should be in the vocab and not the model document; the whole issue is really of relevance for RDF people only, let alone the fact that encoding named graphs may be more readable in TriG rather than JSON-LD.

See you soon!

Ivan


[1] https://www.w3.org/TR/rdf11-concepts/#section-dataset




> Kind regards,
> 
> Antoine
> 


----
Ivan Herman, W3C
Digital Publishing Lead
Home: http://www.w3.org/People/Ivan/
mobile: +31-641044153
ORCID ID: http://orcid.org/0000-0003-0782-2704