[Pellet-users] RDF to DIG 1.1 standalone converter

[Alexander Nazarov]

Hi Mike,

Thank you for your kind advice.

Yes, I'm aware that DIG 1.1 is less expressive than OWL DL, I just hope that
my ontologies would be quite simple.

Yes, the problem is a platform - I was told by my boss that Java is
disfavored in our development. Our domain is NLP and consequently we are in
Perl.

Could you tell me please: what is it, a bridge from Perl to Java?


About converter to DIG 1.1 - I found one called OWL2DIG:
http://projects.semwebcentral.org/projects/owl2dig/

It looks like working OK. The only strange thing I notice:
Every class definition from OWL was converted to <equalc> axiom, i.e. for
class 'Thing':
    <equalc>
        <catom name="http://www.w3.org/2002/07/owl#Thing"/>
        <catom name="http://www.w3.org/2002/07/owl#Thing"/>
    </equalc>

Could you tell me please, is it the same as <defconcept> introduction?

I actually subscribed to DIG forum mailing list:
dig-users at lists.sourceforge.net
but it looks dead.


Thank you and best regards,
Sasha


-----Original Message-----
From: Michael Smith [mailto:msmith at clarkparsia.com] 
Sent: Friday, July 11, 2008 2:57 PM
To: Alexander Nazarov
Cc: pellet-users at lists.owldl.com
Subject: Re: [Pellet-users] RDF to DIG 1.1 standalone converter

On Thu, 2008-07-10 at 12:10 +0100, Alexander Nazarov wrote:

> I have a set of ontologies to hold common domain knowledge.
> Then for every ontology:
> - I have a series of concepts properties extracted from processed
documents.
> - I need to make an inference about those properties by adding them to an
> ontology.
> 
> For a certain reason I can't use OWL API or Jena API to solve my problem.
> I meant to do it by a set of 2 <tell> requests (import KB and import
> properties) and 1 <ask> request (export inference) to Pellet DIG server.
> That's why I need to convert my ontologies from OWL to DIG.
> 
> 
> Thank you very much for your advice about Protege 3.
> Actually I can't find any other means to do it at DIG resources and by
> search in Google.
> That makes me doubt: may be that way to solve my problem is not optimal?

Jena and the OWLAPI have classes to translate from their native
representations to DIG.  Presumably anything that implements a DIG
client will have similar functionality.

Please be aware that the KR supported by DIG 1.1 is less expressive than
OWL DL, so the translation offered by Protege 3, or any other tool, is
likely to lose some information.  The expressivity of DIG 2 will match
OWL 2 DL, but at present there aren't DIG 2 tools.

Without knowing the cause of your OWL API / Jena prohibition, its
difficult for me to suggest other alternatives.
?
If the problem is a platform one (i.e., you're not using Jena and OWLAPI
because you're not in Java), perhaps you can consider using a bridge
from whatever your language is to Java.

-- 
Mike Smith

Clark & Parsia