Re: importing RDF and OWL

>> I added a section [1] to the RDF and OWL compatibility document about 
>> importing RDF/OWL and referring to the entailment regime.
>> [1]
> Nice work.  Here are a few comments (that get especially muddy around
> the promotion of contexts).  Sorry I didn't reply earlier.
> * I can't parse the first sentence or two.  I think there a few words
>   missing.

I updated the text.

> * There's no XML syntax given in BLD for import. What does the XML look
>   like?  What roles names do you propose?

I leave that to the BLD editors.

> * I don't really like the document IRI's being used as Context IRIs.
>   It's useful, but I see it raising some problems when you start to
>   formalize the notion of Contexts.     Let's make up some other IRIs
>   and link them back to the documents.  :-/    Something like

I would be fine with that.  Other opinions?

> * I'm a little concerned about your notion of promotion in a
>   hierarchy of entailment regimes.  If we imagine a world with NAF, then
>   promotion might be incorrect.  Maybe add a flag, so users have to
>   explicitely "import with context = X" or "import with context >= X" ?

When importing RDF/OWL there can be only one context for all imports. 
So, any ruleset with both "import with context= X" and "import with the 
context=Y" be rejected.

Perhaps it would make sense to just specify one context for all imports.

> * What happens if OWL 2 includes a flag (as I hope/expect it will)
>   indicating its own "Context"?  In that (hopefully common) case, should
>   users leave out the context parameter (not possible in the syntax),
>   should they provide a fully generic one, or should whatever they
>   provide be ignored?  That is, is it an OVERRIDE for what the imported
>   document claims to be, or a DEFAULT in case the imported document
>   doesn't say?   I *think* it should be DEFAULT.

According to the current specification, they should provide a generic 
context; then, they are actually free to interpret the combination in 
any way they want.
The current specification actually does not say anything about what to 
do with entailment regimes beyond those specified in the document.  I 
guess we should.

I agree that if the imported document specifies a context, one should 
take that into account (i.e., it should override the specified context).

> * I lean toward having just one Generic context.  Is OWL Full more like
>   RDF or more like OWL?   

I'd say "OWL Full" sounds more like "OWL" than "RDF" :-)

that said, I think it's fine to have just one generic context for RDF/OWL.

>   I can't tell.   I can owl:import some OWL
>   into some RDF, 

No, as soon as you start interpreting that "owl:import" you are in the 
OWL world.

>   or some RDF into some OWL, right?

In that case, you interpret the RDF as OWL.

> * Okay, I guess I'm general confused about whether and how promotion can
>   work....     It sounds like I can't import some RDF triples and an OWL
>   DL ontology at the same time....

If the RDF triples happen to be a valid OWL DL ontology, you can import 
them both as OWL DL.

 >   I'll be forced to treat the DL
>   ontology as OWL-Full, which is unimplementable.

If you start to mix arbitrary RDF with OWL DL, you end up with OWL full, 
yes.  This also happens if you import arbitrary RDF in an OWL DL ontology.

> * Any idea what construct you'd want to use for loading some OO-style
>   XML?  Could we do that with Imports, too?

I guess so.

> * ... I think maybe "Context" can be thought of as "Language".  You're
>   loading some data/rules/knowledge, and naming the language it's
>   written in, in case it's not properly labeled (as RDF data is not).

Yes, that is probably be a better name.
Other opinions?

Best, Jos

> Again, overall it seems quite good, but I don't understand well enough
> how contexts/language/etc is going to work in practice.   
>       -- Sandro

Jos de Bruijn  
An expert is a person who has made all the
mistakes that can be made in a very narrow
   - Niels Bohr

Received on Tuesday, 6 May 2008 08:00:53 UTC