Re: Response to comments of Enrico

I am not quite sure what he wants to say either, but he apologises in
mail 3 that he missed the fact that Simon was talking about OWL RL,
when Simon asked:
How much of OWL 2 RL could be implemented with pure SPARQL.
>From that I guessed that Enrico responds to encoding OWL (or some
fragment thereof) in SPARQL.

hen Enrico says:
If you fix the entailment regime to RDFS, then very little of OWL2 can
be encoded in SPARQL, since it is has been shown that the
computational complexities diverge too much.

Which, is true since a pure RDFS implementation will not really give
you all OWL answers no matter what (RDFS has no notion of
non-determinism for example).

Then Enrico goes on:
There are also simple counter-examples showing that it does not make
sense to have an OWL2 entailment regime in SPARQL, since you would get
unsound results (wrt OWL2 semantics) very easily. So, really, SPARQL
can hardly go beyond RDFS.

Now here it starts to be less clear. Is he still talking about
encoding OWL2 in pure SPARQL? If he is, it makes sense. If not, I can
not quite follow. I can see incomplete results without any
restrictions on the queries and query answers, but he is talking about
unsound results. That makes more sense if he still thinks about
encoding (which is more or less rewriting as I understand it) OWL into
pure SPARQL, i.e., SPARQL queries that are then answered by a system
that uses simple entailment.

Assuming that he is really talking about that, I remove the examples
discussion and instead reply with:

For the OWL entailment regimes we do not envisage an encoding into
SPARQL queries that are then evaluated under simple entailment
semantics. The current working draft includes an OWL Direct Semantics
entailment regime that hopefully clarifies this. The working draft
does not yet define an entailment regime for OWL RDF-Based Semantics
(which is used for the OWL 2 RL profile), but that is envisaged for a
future working draft.


On 27 January 2010 08:32, Ivan Herman <> wrote:
> The sentence
> [[[
> It seems indeed not possible to implement an OWL 2 RL entailment regime
> by using just SPARQL queries at least not with a linear query rewriting
> if that is what you are referring to.
> ]]]
> (though a bit too convoluted:-) made me think.
> The point of the matter is that Enrico does not really state what he is
> thinking about. Do you really believe he referred to implementing OWL 2
> RL (or OWL 2 in general, for the matter) through query rewriting in
> SPARQL? If this is really what his comment is on (I must admit that,
> until I read your sentence, I did not realize that) than all his
> comments are essentially irrelevant, aren't they, because the way OWL is
> approached in this document has nothing to do with query rewriting. If
> this is all correct, than I propose not to even go into the details of
> the discussion on whether his particular entailment is correct or not,
> just tell him that the way we envisage entailment in OWL is not related
> to the query language proper, and that is it...
> Cheers
> Ivan
> On 2010-1-26 23:12 , Birte Glimm wrote:
>> Hi all,
>> as agreed in today's teleconf, I drafted a response to Enrico, which I
>> hope is k for the rest of the group:
>> If you have any comments, please let me know,
>> Birte
> --
> Ivan Herman, W3C Semantic Web Activity Lead
> Home:
> mobile: +31-641044153
> PGP Key:
> FOAF   :
> vCard  :

Dr. Birte Glimm, Room 306
Computing Laboratory
Parks Road
United Kingdom
+44 (0)1865 283529

Received on Wednesday, 27 January 2010 10:04:05 UTC