Christoph; Using SPIN requires a different way of thinking about the
problem. I believe what you want to state is that any use of the a
property needs to have a resource of the correct domain/range type
(correct me if I am misreading your intent). You can state this (more
directly) as a spin:constraint - here's an example for property domain
defs:
# subject does not match defined domain type
ASK WHERE {
?this ?prop ?obj .
?prop rdfs:domain ?domain .
?this a ?domain .
}
The constraint warning will show the text you put in the initial
comment - i.e. all instance of this violation will state "subject does
not match defined domain type (SPIN constraint at owl:Thing)" I
defined this at owl:Thing, but could have defined it for
rdfs:Resource, rdfs:class, or my:fiddlestyx. That's a big modeling
advantage, IMHO.
You can also define this as a CONSTRUCT query, instead of ASK, and
make all violations a member of a class, such as
spin:ConstraintViolation.
Holger's blog (http://composing-the-semantic-web.blogspot.com/) has
excellent examples.
In terms of Pellet 2, if some entity were to do us the favor of
creating a 3rd party plugin, there would be no licensing
complications...
-- Scott
On Feb 9, 11:31 am, "Atanas Kiryakov" <[email protected]> wrote:
> Hello everyone
>
> first: thanks to to everyone who commented on the subject!
>
> Indeed, Ontotext is going no infer inconsistency. There are several things
> to be commented:
>
> - in RDFS semantics there is no notion of inconsistency (at least not at
> this level). Treating global property domain and range definitions (defined
> through rdfs:domain and range) as constrains is *not* the proper
> interpreation. This is for good reason RDFS is defined to support open-world
> semantics in non-controlled web environment. There are obviously scenarios
> where inconsistency interpretation is necessary (mostly in DB-style
> closed-world setups). Such constrainst are available in other languages
> (e.g. some flavors of OWL 2, as well as WSML,http://www.wsml.org/)
>
> - ter Horst, [1], defines two inconsistency rules which are easy to
> interpret with datalog-type inference "...
> - a combination of two triples of the form v differentFrom w, v sameAs w, or
> - a combination of three triples of the form v disjointWith w, u type v, u
> type w. ..."
>
> - OWLIM can easily support the above type of incompleteness inference, but
> there were two minor obstacles: (i) Sesame framework does not provide a
> standard mechanism to report such inconsistencies and (ii) we need a minor
> alternation in the rule language of OWLIM, so, that such rules can be
> defined and .
>
> - will provide inconsistency support in one of the next versions of OWLIM.
> We will solve the minor "infrastructure" issues (rule language and APIs) and
> provide rulesets (inference profiles) where: (i) the terHorst inconsistency
> rules are added and (ii) alternative domain and range support where those
> are treated as inconsistency check, instead of generative inference rule
>
> Meanwhile, one can check for inconsistencies by wrapping the above-mentioned
> inconsistency rules as queries.
>
> Regards,
> Naso
>
> [1]http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B758F-4H16P4...
>
> ----------------------------------------------------------
> Atanas Kiryakov
> CEO of Ontotext AD,http://www.ontotext.com
> Sirma Group Corp,http://www.sirma.bg
> Phone: (+359 2) 8091 555; Fax: 8090 404
> ----------------------------------------------------------
>
> ----- Original Message -----
> From: "Dave McComb" <[email protected]>
> To: <[email protected]>
> Sent: Monday, February 09, 2009 6:21 PM
> Subject: [tbc-users] Re: owlim
>
> Yeah, the issue is OWLIM doesn't do disjoints (or complements), it does the
> domain and range fine but won't enforce the disjointness
>
> > -----Original Message-----
> > From: [email protected] [mailto:topbraid-
> > [email protected]] On Behalf Of Christoph
> > Sent: Monday, February 09, 2009 8:23 AM
> > To: TopBraid Composer Users
> > Subject: [tbc-users] Re: owlim
>
> > but if I have another concept called eType, that is disjoint with
> > dType and I have the instance e1 of Type eType:
> > and define the triple e1 :aProp :r2
> > then the reasoner should complain about this as e1 cannot be of Type
> > dType, but OWLIM does not.
>
> > Thanks, Christoph
>
> > On 9 Feb., 16:07, Scott Henninger <[email protected]> wrote:
> > > Christoph; The inferences you can make with domain and range are
> > that
> > > given domain and/or range restrictions on a property, you can infer
> > > the type of a resource. For example, given the triples:
> > > :r1 :aProp :r2
> > > :aProp rdfs:domain :dType
> > > :aProp rdfs:range :rType
>
> > > The following inference will be made:
> > > :r1 rdf:type :dType
> > > :r2 rdf:type :rType
>
> > > SwiftOWLIM will make this inference. BTW, there is an excellent
> > > treatment of this and other RDFS/OWL inferences in Allemang &
> > > Handler's Working Ontologist book. Highly recommended!
>
> > > -- Scott
>
> > > On Feb 9, 7:56 am, Christoph <[email protected]> wrote:
>
> > > > If I have a property with domain and range restriction and I use
> > this
> > > > property with another domain (both domains are disjoint) then there
> > is
> > > > no error when reasoning is applied.
>
> > > > Does OWLIM not support domain/range restrictions?
>
> > > > br, Christoph- Zitierten Text ausblenden -
>
> > > - Zitierten Text anzeigen -
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups
"TopBraid Composer Users" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to
[email protected]
For more options, visit this group at
http://groups.google.com/group/topbraid-composer-users?hl=en
-~----------~----~----~----~------~----~------~--~---
