Re: relevant paper on using OWL (description logics) for RDF validation

On 10/25/2014 04:01 PM, Holger Knublauch wrote:
>
> On 10/26/14, 8:52 AM, Peter F. Patel-Schneider wrote:
>> I'm not sure how this would work.  Are you saying that taking an arbitrary
>> RDF graph containing some OWL axioms and adding the right imports triple
>> would be all that is needed to pass the graph into a SPIN processor and get
>> the behaviour you state above?  This doesn't seem possible.
>
> Why not? SPIN doesn't care about inferencing.

But then what is all the talk about SPIN working with OWL (or RDFS, for that 
matter)?

In particular, the results of SPIN rules should participate in whatever other 
inferencing is going on.

>If someone wants to run
> constraint checks over an OWL model then it needs to run the SPARQL processor
> over a graph that has the additional inferences visible.

Well, the situation is not nearly so simple as this.  The additional 
inferences may be infinite, for example.

> The same applies for
> any other inferencing mechanism such as a rule language.

Well that depends on the power of the rule language.  Datalog rules, for 
example, are often unproblematic.  Adding in even just functions, can easily 
change the situation dramatically.

>> In any case, this doesn't seem to be the way that constraint validation
>> should work.  You need at least three kinds of information - data, axioms,
>> and constraints - and this appears to provide only two of them.
> Why? Which aspect is missing?

Well, I thought that there were supposed to be three separate kinds of 
information.  If all that SPIN sees is the results of inference, then the 
three kinds of information are no longer separate.

Maybe SPIN doesn't need the information separated out.  However, then 
inference can produce new spin:constraint relationships.

[...]

> Holger

peter

Received on Sunday, 26 October 2014 00:03:51 UTC