Re: rdf:nodeID and rdf:about in powder-dr examples

Kanzaki san,

Thank you very much for this very helpful mail. You have pointed out 
something none of us spotted (a case of being too close to it I think, 
hence the joy of publication and review!).

The solution for us is to use nodeID throughout since it is very 
deliberate that you cannot add further triples to an IRI set or 
descriptor set from outside (which all comes from POWDER's obsession 
with data provenance).

And we'll fix the typos in 2-7.

Thanks again,

Phil.

KANZAKI Masahide wrote:
> Hi,
> 
> In POWDER-S examples of powder-dr document, all classes that define
> IRI-set/Descriptor-set are denoted by rdf:nodeID, while class
> descriptions that add additional information to those sets are denoted
> by rdf:about/rdf:resource. eg:
> 
> <owl:Class rdf:nodeID="iriset_1">
> ...
> </owl:Class>
> <owl:Class rdf:nodeID="descriptorset_1">
> ...
> </owl:Class>
> <owl:Class rdf:about="#iriset_1">
>     <rdfs:subClassOf rdf:resource="#descriptorset_1"/>
> </owl:Class>
> 
> This does not make any meaningful relationship between those sets. I
> wonder, set definitions should be denoted by rdf:ID (not nodeID), or
> additional descriptions should refer those sets with rdf:nodeID (not
> rdf:about/rdf:resource). In case of reference across different files,
> rdf:nodeID doesn't work.
> 
> Please make sure that rdf:nodeID is used for a blank node, while
> rdf:about/rdf:resource is for a node with URI.
> 
> (and, I guess "#description_1" etc in Example 2-7 are typo of
> "#descriptorset_1" ...)
> 
> best regards,
> 

-- 
Phil Archer
Chief Technical Officer,
Family Online Safety Institute
w. http://www.fosi.org/people/philarcher/

Received on Thursday, 3 July 2008 08:40:22 UTC