RE: How to handle predicates and anonymous inferred classes created in protege via JENA?

Paul Gearon wrote:

>Hi,
>
>Discussion at the end...
>
>On Mon, Jun 21, 2010 at 2:54 PM, mihir sanghavi <ms478@njit.edu> wrote:
>> ---------- Forwarded message ----------
>> From: <fa2260@columbia.edu>
>> Date: Mon, Jun 21, 2010 at 5:38 PM
>> Subject: Re: How to handle predicates and anonymous inferred classes
>created
>> in protege via JENA?
>> To: mihir sanghavi <ms478@njit.edu>
>>
>>
>> Hey Folks,
>>
>> I am developing a semantic web application that helps navigate
>biological
>> hierarchies. The ontology is modelled in protege. I am having problems
>> parsing the Protege generated .owl file.
>>
>> As an example. I modeled the relationship, "AdventititiousRoot
>isPartOf only
>> Root" (isPartOf is a user-defined porperty). I modeled/described this
>> relationship under the 'SuperClass' field for the Class
>'AdeventitiousRoot'.
>> The 'Class Usage' generated was "AdventitiousRoot subClassOf isParOf
>only
>> Root". The associated rdf/xml code is given below.
>>
>> <!--
>>
>http://www.semanticweb.org/ontologies/2010/1/3/PlantEntities.owl#Adventi
>tousRoot
>> -->
>>
>>    <owl:Class rdf:about="#AdventitousRoot">
>>        <rdfs:subClassOf rdf:resource="&owl;Thing"/>
>>        <rdfs:subClassOf>
>>            <owl:Restriction>
>>                <owl:onProperty rdf:resource="#isPartOf"/>
>>                <owl:allValuesFrom rdf:resource="#Root"/>
>>            </owl:Restriction>
>>        </rdfs:subClassOf>
>>    </owl:Class>
>>
>> I parsed this code using JENA's object "statement iterator" and I got
>> triples such as "AdventitiiousRoot subClassOf Thing" and
>"AdventitiousRoot
>> subclassOf 6b8720b3:128d06ed301:-7fc6".
>>
>> I am able to capture the Subject, Predicate and Object in the first
>> statement.
>>
>> My problem is how do I capture the predicate "isPartOf" and the owl
>> restriction "only" and avoid generating the hexadecimal reference in
>the
>> second statement?
>
>(I'll follow your lead, and skip the domain prefixes for clarity)
>
>I think you need to understand the RDF structure that you have here.
>
>The RDF/XML above is describing AdventitousRoot as a subClassOf the
>resource called Thing, and also as a subClass of another resource that
>is a Restriction. This Restriction resource has no name, and hence it
>is represented with a blank node. Identifier or not, the system needs
>to have some kind of internal representation to uniquely identify this
>resource, so it has generated one. This is the reason for the
>hexadecimal label that you see.
>
>The unnamed Restriction resource has two further properties which
>describe how it does restrictions. It uses the onProperty property to
>describe that it restricts the isPartOf predicate, and it uses the
>allValuesFrom property is restrict isPartOf to just refer to resources
>of type Root.
>
>Everything here is perfectly correct RDF representation of an OWL
>class. Apparently, you don't like the fact that your Restriction class
>has no identifier. The way around it would be to give it a name.
>That's valid RDF, though I can't say if there's a problem in OWL (can
>someone enlighten me please? I *think* it's fine, right?). 

It depends on what you mean by a "name" here, and on what you mean by "OWL"
here :). 

In the RDF serialization of OWL (2) DL, the "root node" of a property
restriction is always a bNode, see Table 13 in [1]. If you put there a URI
instead, then the mapping is formally undefined. I don't know how actual
tools handle such cases, though, but parsing may fail. Anyway, the standard
way to provide a "name" for a property restriction is via a class
equivalence axiom:

    ex:MyRestriction owl:equivalentClass _:x .
    _:x rdf:type owl:Restriction .
    _:x owl:onProperty ex:p .
    _:x owl:allValuesFrom ex:C .

In OWL Full, you can use a URI instead of the bNode. Or even, if you like,
associate a URI with the bNode via owl:sameAs.

Michael

[1] <http://www.w3.org/TR/2009/REC-owl2-mapping-to-rdf-20091027/>

>Also,
>Protege may not deal with the name correctly (it may ignore it... I
>don't know).
>
>Anyway, the provide the Restriction with a name, you only need to add
>an rdf:about. The following uses the name AdventitousRestriction in
>the same domain as AdventitousRoot:
>
>
><owl:Class rdf:about="#AdventitousRoot">
>  <rdfs:subClassOf rdf:resource="&owl;Thing"/>
>  <rdfs:subClassOf>
>    <owl:Restriction rdf:about="#AdventitousRestriction">
>      <owl:onProperty rdf:resource="#isPartOf"/>
>      <owl:allValuesFrom rdf:resource="#Root"/>
>    </owl:Restriction>
>  </rdfs:subClassOf>
></owl:Class>
>
>
>Regards,
>Paul Gearon

--
Dipl.-Inform. Michael Schneider
Research Scientist, Information Process Engineering (IPE)
Tel  : +49-721-9654-726
Fax  : +49-721-9654-727
Email: michael.schneider@fzi.de
WWW  : http://www.fzi.de/michael.schneider
=======================================================================
FZI Forschungszentrum Informatik an der Universität Karlsruhe
Haid-und-Neu-Str. 10-14, D-76131 Karlsruhe
Tel.: +49-721-9654-0, Fax: +49-721-9654-959
Stiftung des bürgerlichen Rechts, Az 14-0563.1, RP Karlsruhe
Vorstand: Prof. Dr.-Ing. Rüdiger Dillmann, Dipl. Wi.-Ing. Michael Flor,
Prof. Dr. Dr. h.c. Wolffried Stucky, Prof. Dr. Rudi Studer
Vorsitzender des Kuratoriums: Ministerialdirigent Günther Leßnerkraus
=======================================================================

Received on Tuesday, 22 June 2010 16:25:31 UTC