Re: Question about dataranges incompatibilities

What does it matter that we don't have such an axiom? It's still true. Indeed in the structural specification you can see such language in eg section 4.2:

"As specified in XML Schema [XML Schema Datatypes<http://www.w3.org/TR/owl2-syntax/#ref-xml-schema-datatypes>], the value spaces ofxsd:double, xsd:float, and xsd:decimal are pairwise disjoint."

You can't express disjointness as an axiom, but for good reason. There's no real case where this makes sense.

We always know the members of a data range as well as the identity conditions of its members. So we can always derive the disjointness. If we asserted a disjointness, then the most we could do was contradict the expressions. That doesn't seem all that useful. We also can't express subsumptions though we know they exist.

A query language might add such tests and certainly an explanation tool could make good use of such.

On Dec 21, 2014, at 9:08, "Leila Bayoudhi" <bayoudhileila@yahoo.fr<mailto:bayoudhileila@yahoo.fr>> wrote:

Yes, this is what I mentioned for non datatype cases since we have already what we call disjointClassesAxiom(A, B)
But for datatype we have not explicit axiom defing the disjoitness relation between two data ranges.


Le Dimanche 21 décembre 2014 14h36, Bijan Parsia <bijan.parsia@manchester.ac.uk<mailto:bijan.parsia@manchester.ac.uk>> a écrit :


I would say that the data types are disjoint thus their intersection of value sets is the empty set.

This is strictly analogous to the non datatype case:

A disjointwith B.
C = P someValuesFrom A and
       P allValuesFrom B.

On Dec 21, 2014, at 7:14, "Leila Bayoudhi" <bayoudhileila@yahoo.fr<mailto:bayoudhileila@yahoo.fr>> wrote:

Hi,
Given  new defined datatype called newdata
 <DatatypeDefinition>
        <Datatype IRI="#newData"/>
        <DatatypeRestriction>
            <Datatype abbreviatedIRI="xsd:int"/>
            <FacetRestriction facet="&xsd;maxInclusive">
                <Literal datatypeIRI="&xsd;integer">100</Literal>
            </FacetRestriction>
            <FacetRestriction facet="&xsd;minExclusive">
                <Literal datatypeIRI="&xsd;integer">0</Literal>
            </FacetRestriction>
        </DatatypeRestriction>
    </DatatypeDefinition>

Given  a subclassOf axiom
<SubClassOf>
        <Class IRI="#person"/>
        <DataAllValuesFrom>
            <DataProperty IRI="#hasIdentifier"/>
            <Datatype IRI="#newData"/>
        </DataAllValuesFrom>
    </SubClassOf>
Intentionally for making unsatisfiable the class person, I introduce this subClassOf axiuom
<SubClassOf>
        <Class IRI="#person"/>
        <DataSomeValuesFrom>
            <DataProperty IRI="#hasIdentifier"/>
            <Datatype abbreviatedIRI="xsd:boolean"/>
        </DataSomeValuesFrom>
    </SubClassOf>
    <SubClassOf>

here what can i to explain "formally" this unsatisfiability:( concerning -DR1:= newdata DR2:= boolean)
can we say DR1 DR2 are different?
can we say DR1 DR2 are incompatible?
can we say that the value space of DR1 IS NOT in the DR2 one?
or what we say exactly?

Thank you for helping me.