Page 1 of 1
ClassifiedIdentificationOfClass
Posted: Wed Feb 08, 2012 10:54 am
by HansTeijgeler
- ClassifiedIdentificationOfClass.png (19.02 KiB) Viewed 10532 times
I propose the following:
Name: ClassifiedIdentificationOfClass
Intent: Identification of a class with a string, according a given type of identification (e.g. RDL Unique Name).
Description: ClassifiedIdentificationOfClass(x1, x2, x3) means that x1 is a [class], x2 is a string and x3 a [class_of_class_of_identification], and that x2 is an x3-type identification of x1.
Graph ClassifiedIdentificationOfClass.png is attached.
Code: Select all
Signature:
1 hasIdentified http://dm.rdlfacade.org/data#Class &rdl;R12347255217
2 valIdentifier http://dm.rdlfacade.org/data#ExpressString EjectorPump
3 hasIdentificationType http://dm.rdlfacade.org/data#ClassOfClassOfIdentification &rdl;R3256432
Re: ClassifiedIdentificationOfClass
Posted: Wed Feb 08, 2012 2:54 pm
by OnnoPaap
The name of a class is a special case.
In the example the classification of the identification is Unique Name. This is in fact an annotation property in ISO 15926-8, annUniqueName, subProperty of rdfs:label, which can also be equipped with the lang= attribute.
In that case, do we still use the template?
Re: ClassifiedIdentificationOfClass
Posted: Wed Feb 08, 2012 3:44 pm
by HansTeijgeler
If I had used any other example, like ASME Code or the like, then it would be OK with you?
But having used the example of RDL Unique Name is still useful, because it raises the question what to do when subRDLs are being produced in another language?
Are we then editing the RDL records by adding annUniqueName's in those languages? Seems odd from a secutity point of view.
With this template, or rather one which is extended with the information in which language the name is, the Russians, Koreans, French, Germans, Chinese etc can build their own set of names in their native languages, but at the same time maintaining the link with the logic of the RDL.
I will propose the extended template, named ClassifiedIdentificationOfClassInlanguage, separately.
Re: ClassifiedIdentificationOfClass
Posted: Tue Feb 28, 2012 10:35 am
by vvagr
We've to extend the format to have an ability to state that this template is a specialization of ClassifiedIdentification template.
We're working on search language for ISO 15926 data, and we've very clear understanding now - use of template specialization greatly enhances our ability to work with common patterns of data representation.
Re: ClassifiedIdentificationOfClass
Posted: Tue Feb 28, 2012 1:01 pm
by HansTeijgeler
Hi Victor,
A TemplateStatement (part 8) is a combination of ClassOfInformationRepresentation AND MultidimensionalObject. Part 2 allows that combination since they are not in a ONEOF subtype relation with AbstractObject.
So, yes, specialization is possible in the OWL code, but that is not what you have in mind, since the proposal does not show OWL code. In the TemplateSpecification we can include that. I adapted the example TS accordingly, see
http://www.infowebml.ws/TS/ClassifiedId ... fClass.xml
Did I miss somethng?
Regards,
Hans
Re: ClassifiedIdentificationOfClass
Posted: Tue Feb 28, 2012 3:39 pm
by OnnoPaap
OnnoPaap wrote:The name of a class is a special case.
In the example the classification of the identification is Unique Name. This is in fact an annotation property in ISO 15926-8, annUniqueName, subProperty of rdfs:label, which can also be equipped with the lang= attribute.
In that case, do we still use the template?
Hans and I discussed this. I now think we should have the template for the unique name in addition to the annotation property annUniqueName (subProperty of rdfs:label).
The unique name is a special case, as it has to be available for Semantic Web tools which need to have the rdfs:label available for them to operate. However a plant item (or activity, or anything) has many different names for all of which the above template must be used. So I think the template must be available also for the unique name, in order to have a uniform way of searching and finding names.