W3C home > Mailing lists > Public > public-rif-wg@w3.org > January 2010

Re: Importing RDF documents from RIF - part 1

From: Axel Polleres <axel.polleres@deri.org>
Date: Tue, 19 Jan 2010 15:07:03 +0000
Cc: <public-rif-wg@w3.org>
Message-Id: <84B6605F-8B67-4AD9-B6FD-95EBC7CBCAC2@deri.org>
To: Dave Reynolds <dave.e.reynolds@googlemail.com>

On 19 Jan 2010, at 14:37, Dave Reynolds wrote:

> The original comment was about importing RDF into RIF, you are talking
> the about the reverse and I don't think any of this relates directly to
> the public comment.
> I, for one, don't understand the motivation of what you are trying to do
> here and so find it hard to comment on the options. What is the value if
> specifying import of RIF into RDF when we have no RDF representation of
> RIF?
> How would this help SPARQL? I would have thought what SPARQL needs is a
> notion of an entailment regime that can be specified by a RIF document.
> That seems rather different from importing a RIF document into an RDF graph.
> Dave

For one, true, that's not the public comments concern, I guess just the subject line made me jump on that topic again.
For the latter... why do I need that for SPARQL:
 SPARQL BGP matching extensions are defined in terms of RDF graphs, not in terms of RIF rulesets. There is no simple 
way to define a RIF semantics for SPARQL (at least I couldn't think of any) without having a pointer from a graph 
to a RIF document. The extension I suggested in the second mail seems to me small enough (but still probably too 
late to get in in the documents... I would be happy enough to suggest something like this in a WG note which we 
could later pick up in the SPARQL WG.) ... Summarizing, such an import mechanism would make life much easier for me :-)
plus I think it is quite useful.

Received on Tuesday, 19 January 2010 15:07:39 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:47:57 UTC