W3C home > Mailing lists > Public > public-prov-wg@w3.org > February 2012

Re: Update on PROV-O OWL file (Action item 55)

From: Stian Soiland-Reyes <soiland-reyes@cs.manchester.ac.uk>
Date: Thu, 16 Feb 2012 12:21:02 +0000
Message-ID: <CAPRnXt=yLCHnPROf1UxZ3Tg-0XK+GSL0ir8B0CvQ1vgP65zzxQ@mail.gmail.com>
To: Ivan Herman <ivan@w3.org>
Cc: Satya Sahoo <satya.sahoo@case.edu>, Provenance Working Group WG <public-prov-wg@w3.org>
That is good feedback. I don't think we have big trouble addressing either.
prov:Time was defined like that when it was time:TemporalEntity, we don't
need that restriction.

Similarly for hadTemporalValue, we can just keep the "scruffy" subproperies
alone + one new for QualifiedInvolvement. Alternatively we can keep the
property and let its domain be blank (owl:Thing).
On Feb 16, 2012 9:24 AM, "Ivan Herman" <ivan@w3.org> wrote:

> Satya et al,
>
> My apologies for chiming in from the 'outside'...
>
> I am not a logician, so you may want to check this with a real expert. But
> in my reading of the OWL 2 Profile document[1], this OWL ontology does not
> fall into OWL 2 RL[2]. If that is true, that may be a problem; it may
> jeopardize its acceptance by large communities that do not think in terms
> of more complex OWL reasoners but which do begin to adopt OWL 2 RL.
>
> The issue I have is with:
>
> <owl:ObjectProperty rdf:about="hadTemporalValue">
>    <rdf:type rdf:resource="&owl;IrreflexiveProperty"/>
>    <rdfs:label xml:lang="en">has temporal value</rdfs:label>
>    <rdfs:range rdf:resource="Time"/>
>    <rdfs:domain>
>       <owl:Class>
>            <owl:unionOf rdf:parseType="Collection">
>                <rdf:Description rdf:about="Activity"/>
>                <rdf:Description rdf:about="Entity"/>
>                <rdf:Description rdf:about="QualifiedInvolvement"/>
>            </owl:unionOf>
>        </owl:Class>
>    </rdfs:domain>
> </owl:ObjectProperty>
>
> Indeed, if I look at the formal spec of the OWL 2 RL document, it says:
>
> ObjectPropertyDomain := 'ObjectPropertyDomain' '(' axiomAnnotations
> ObjectPropertyExpression superClassExpression ')'
>
> with
>
> superClassExpression :=
>    Class other than owl:Thing |
>    superObjectIntersectionOf | superObjectComplementOf |
>    superObjectAllValuesFrom | ObjectHasValue | superObjectMaxCardinality |
>    DataAllValuesFrom | DataHasValue | superDataMaxCardinality
> superObjectIntersectionOf := 'ObjectIntersectionOf' '('
> superClassExpression superClassExpression { superClassExpression } ')'
> superObjectComplementOf := 'ObjectComplementOf' '(' subClassExpression ')'
> superObjectAllValuesFrom := 'ObjectAllValuesFrom' '('
> ObjectPropertyExpression superClassExpression ')'
> superObjectMaxCardinality :=
>    'ObjectMaxCardinality' '(' zeroOrOne ObjectPropertyExpression [
> subClassExpression ] ')' |
>    'ObjectMaxCardinality' '(' zeroOrOne ObjectPropertyExpression owl:Thing
> ')'
> superDataMaxCardinality := 'DataMaxCardinality' '(' zeroOrOne
> DataPropertyExpression [ DataRange ] ')'
>
> ie, using a union of classes as part of the domain is not allowed. The
> rules also express this. And, although a layperson in terms of hard core
> logic, I can see why: if a resource is the subject of that property, a
> simple rule engine _cannot_ find out which of the constituents of the union
> it belongs to. Ie, it cannot make any intelligent deduction.
>
> A similar issue arises with:
>
> <owl:Class rdf:about="Time">
>    <owl:equivalentClass>
>        <owl:Class>
>            <owl:unionOf rdf:parseType="Collection">
>                <rdf:Description rdf:about="Instant"/>
>                <rdf:Description rdf:about="Interval"/>
>            </owl:unionOf>
>        </owl:Class>
>    </owl:equivalentClass>
>    <rdfs:comment rdf:datatype="&xsd;string"
>        >TemporalEntity represents time information, both time instant
> (having 0 time duration) and interval (having a non-zero
> duration).</rdfs:comment>
> </owl:Class>
>
> The corresponding functional spec for OWL 2 RL is indeed:
>
> EquivalentClasses := 'EquivalentClasses' '(' axiomAnnotations
> equivClassExpression equivClassExpression { equivClassExpression } ')'
>
> equivClassExpression :=
>    Class other than owl:Thing |
>    equivObjectIntersectionOf |
>    ObjectHasValue |
>    DataHasValue
> equivObjectIntersectionOf := 'ObjectIntersectionOf' '('
> equivClassExpression equivClassExpression { equivClassExpression } ')'
>
> I.e., again, we do not have a Union.
>
> I am not sure what this means back in the model that is transcribed into
> OWL.
>
> Ivan
>
>
> [1] http://www.w3.org/TR/owl2-profiles/
> [2] http://www.w3.org/TR/owl2-profiles/#OWL_2_RL
>
> On Feb 14, 2012, at 23:24 , Satya Sahoo wrote:
>
> > Hi all,
> > After two meetings by the PROV-O team on Feb 13 and Feb 14 [1], we have
> updated the OWL file and made it available for review by the WG [2].
> >
> > We would like to note that we support only the wasStartedBy defined
> between an activity and an agent, and not wasStartedBy defined between
> activity and activity, and activity and entity (the definition of
> wasStartedBy is defined in three ways by DM-TPWD in Section 6.2 and Section
> 5.3.2.2).
> >
> > All feedback is welcome!
> >
> > Thanks.
> >
> > Best,
> > Satya
> >
> > [1] Minutes of the PROV-O telcon:
> http://www.w3.org/2011/prov/wiki/PIL_OWL_Ontology_Meeting_2012-02-13
> > [2] PROV-O OWL File:
> http://dvcs.w3.org/hg/prov/raw-file/default/ontology/ProvenanceOntology.owl
>
>
> ----
> Ivan Herman, W3C Semantic Web Activity Lead
> Home: http://www.w3.org/People/Ivan/
> mobile: +31-641044153
> FOAF: http://www.ivan-herman.net/foaf.rdf
>
>
>
>
>
>
Received on Thursday, 16 February 2012 12:21:35 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 26 April 2012 13:06:56 GMT