It's not clear exactly what you mean by subproperties. It's important to realize that RDF is a graph-based data representation based on labeled, directed, edges called triples of the form:
subject predicate object
Your data for instance, includes the triples
DEP_05 rdf:type geo:Department
DEP_05 geo:code_department "05"
DEP_05 subdivision ARR_O51
ARR_051 rdf:type geo:Arrondissement
ARR_051 geo:code_arrondissement "051"
RDF/XML provides lots of different ways to write the same graph. For instance, one way of writing your data (I've added the appropriate prefixes) is:
<rdf:RDF
xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
xmlns:geo="https://stackoverflow.com/q/21383685/1281433/">
<geo:Departement rdf:about="https://stackoverflow.com/q/21383685/1281433/DEP_05">
<geo:code_departement>05</geo:code_departement>
<geo:subdivision>
<geo:Arrondissement rdf:about="https://stackoverflow.com/q/21383685/1281433/ARR_051">
<geo:code_arrondissement>051</geo:code_arrondissement>
<geo:nom xml:lang="fr">Briançon</geo:nom>
</geo:Arrondissement>
</geo:subdivision>
<geo:subdivision>
<geo:Arrondissement rdf:about="https://stackoverflow.com/q/21383685/1281433/ARR_052">
<geo:code_arrondissement>052</geo:code_arrondissement>
<geo:nom xml:lang="fr">Gap</geo:nom>
</geo:Arrondissement>
</geo:subdivision>
</geo:Departement>
</rdf:RDF>
Another way, which happens to use fewer of the "shortcuts" that RDF/XML allows, is:
<rdf:RDF
xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
xmlns:geo="https://stackoverflow.com/q/21383685/1281433/" >
<rdf:Description rdf:about="https://stackoverflow.com/q/21383685/1281433/ARR_052">
<rdf:type rdf:resource="https://stackoverflow.com/q/21383685/1281433/Arrondissement"/>
<geo:code_arrondissement>052</geo:code_arrondissement>
<geo:nom xml:lang="fr">Gap</geo:nom>
</rdf:Description>
<rdf:Description rdf:about="https://stackoverflow.com/q/21383685/1281433/ARR_051">
<rdf:type rdf:resource="https://stackoverflow.com/q/21383685/1281433/Arrondissement"/>
<geo:code_arrondissement>051</geo:code_arrondissement>
<geo:nom xml:lang="fr">Briançon</geo:nom>
</rdf:Description>
<rdf:Description rdf:about="https://stackoverflow.com/q/21383685/1281433/DEP_05">
<rdf:type rdf:resource="https://stackoverflow.com/q/21383685/1281433/Departement"/>
<geo:code_departement>05</geo:code_departement>
<geo:subdivision rdf:resource="https://stackoverflow.com/q/21383685/1281433/ARR_051"/>
<geo:subdivision rdf:resource="https://stackoverflow.com/q/21383685/1281433/ARR_052"/>
</rdf:Description>
</rdf:RDF>
Even though these are different XML documents, they are serializations of the same RDF graph. There are non-XML serializations as well. In the Turtle serialization, your data is:
@prefix geo: <https://stackoverflow.com/q/21383685/1281433/> .
@prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#> .
geo:ARR_051 a geo:Arrondissement ;
geo:code_arrondissement "051" ;
geo:nom "Briançon"@fr .
geo:DEP_05 a geo:Departement ;
geo:code_departement "05" ;
geo:subdivision geo:ARR_051 , geo:ARR_052 .
geo:ARR_052 a geo:Arrondissement ;
geo:code_arrondissement "052" ;
geo:nom "Gap"@fr .
It's very important to realize that these are all the same graph. An RDF processing tool doesn't care which one you use. That said, it can be nice to use a format that's more human-readable, so I tend to prefer Turtle. If you still need to use RDF/XML, then you may way to serialize your model using the "RDF/XML-ABBREV" language. To do this, specify "RDF/XML-ABBREV" as the lang argument to Model.write(OutputStream out, String lang). From the Javadoc:
Model write(OutputStream out, String lang)
Write a serialized represention of this model in a specified language.
The language in which to write the model is specified by the lang
argument. Predefined values are "RDF/XML", "RDF/XML-ABBREV",
"N-TRIPLE", "TURTLE", (and "TTL") and "N3". The default value,
represented by null, is "RDF/XML".
Parameters:
- out - The output stream to which the RDF is written
- lang - The output language
Returns:
Related Questions
Some related questions have arisen here before, and you might find some of them helpful. The ones that I quickly found are:
