Victor Agroskin and myself have been seeking for a simplification of the templates for Class and for temporal parts, thereby trying to make the signatures of templates for individual and for temporal part identical.
The big advantage of that would be that the templates for individual, produced by the IIP group and others, would automatically also become suitable for information integration.
The solution found by us is:
* The classOfPart Class and the temporal part Individual are no longer explicitly listed in the signature
* For each template the use of meta:valEffectiveDate and meta:valDeprecationDate is mandatory; this covers, for the lowered templates in RDF, the timestamping requirements
* For templates for temporal part this would lead to two predicates with essential the same information: rdl:valStartTime and meta:valEffectiveDate
* For SPARQL queries it is convenient to use meta:valEffectiveDate for ALL templates; hence the omission of rdl:valStartTime from the signatures of templates for temporal part
* In cases where OWL reasoning is required a subset of the lifecycle information is selected and subsequently mapped to lifted templates in OWL format, thus "unhiding" the classOfPart Classes and the temporal part Individuals.
* The time aspect of the lifted templates for Class shall be taken care of by using the lifted templates ActivityCausesEffectiveClass (http://www.15926.org/templatespecs/CL-PTYST-08.xml) and ActivityCausesDeprecatedClass (http://www.15926.org/templatespecs/CL-PTYST-09.xml), thereby addressing the applicable classOfPart Classes
* The time aspect of the lifted templates for Individual shall be taken care of by using the value of meta:valEffectiveDate for the RepresentationOfGregorianDateAndUtcTime.
EXAMPLES
A small set has been prepared to demonstrate some of the above.
Please look at http://www.15926.org/templatespecs/CL-IDENT-02.xml and http://www.15926.org/templatespecs/IN-IDENT-02.xml to see examples of the changed template style as discussed above.
All temporal part templates such as http://www.15926.org/templatespecs/TP-IDENT-02.xml would disappear (they are now under the guise of templates for individual, see above).
The templates for WholeLifeIndividual, like http://www.15926.org/templatespecs/WI-IDENT-02.xml will stay.
Please comment
Hiding the (class_of)temporal part in template signatures
-
- Posts: 283
- Joined: Sun Jan 22, 2012 10:02 pm
-
- Posts: 77
- Joined: Tue May 15, 2012 8:48 am
Re: Hiding the (class_of)temporal part in template signature
This is very interesting Hans
How would the meta:valDeprecationDate be used ?
Keith
How would the meta:valDeprecationDate be used ?
Keith
Re: Hiding the (class_of)temporal part in template signature
I know you won't use e.g. meta:valEffectiveDate as a role in the template signature to make the signature of possible individual templates identical to the temporal parts templates.
But shouldn't e.g. meta:valEffectiveDate be noted in the diagram at least?
But shouldn't e.g. meta:valEffectiveDate be noted in the diagram at least?
-
- Posts: 283
- Joined: Sun Jan 22, 2012 10:02 pm
Re: Hiding the (class_of)temporal part in template signature
@Keith -
For templates for Individual use the template EndingOfIndividual http://www.15926.org/templatespecs/IN-EXIST-03.xml where the individual is the "unhidden" temporal part Individual in the template that is deprecated, and the value of meta:valDeprecationDate is used for the RepresentationOfGregorianDateAndUtcTime.
For templates for Class as indicated: use template ActivityCausesDeprecatedClassHow would the meta:valDeprecationDate be used ?
For templates for Individual use the template EndingOfIndividual http://www.15926.org/templatespecs/IN-EXIST-03.xml where the individual is the "unhidden" temporal part Individual in the template that is deprecated, and the value of meta:valDeprecationDate is used for the RepresentationOfGregorianDateAndUtcTime.
Re: Hiding the (class_of)temporal part in template signature
We can try to extend the concept of pattern to include such thing as annotation role which is optional and is included into templates when the pattern is expanded.
-
- Posts: 283
- Joined: Sun Jan 22, 2012 10:02 pm
Re: Hiding the (class_of)temporal part in template signature
@Onno - We could do that, but it cannot become a part of the FOL code, because the lowered template disappears in the OWL representation of the lifted template. It would look like this:
It can be done, but it can also be done in the notes. Just tell me.[/size]
-
- Posts: 283
- Joined: Sun Jan 22, 2012 10:02 pm
Re: Hiding the (class_of)temporal part in template signature
@ Victor - We first have to come to terms about what a pattern exactly is, and how that should be implemented.
I am working with Onno on a proposal, due this week.
I am working with Onno on a proposal, due this week.
-
- Posts: 77
- Joined: Tue May 15, 2012 8:48 am
Re: Hiding the (class_of)temporal part in template signature
Ok Perhaps I was not specific enough.HansTeijgeler wrote:@Keith -For templates for Class as indicated: use template ActivityCausesDeprecatedClassHow would the meta:valDeprecationDate be used ?
For templates for Individual use the template EndingOfIndividual http://www.15926.org/templatespecs/IN-EXIST-03.xml where the individual is the "unhidden" temporal part Individual in the template that is deprecated, and the value of meta:valDeprecationDate is used for the RepresentationOfGregorianDateAndUtcTime.
In your initial message you made the statement
So lets say I create an instance of the template ClassifiedIdentificationOfClassOfIndividual I can see easily enough what I set the annotation meta:valEffectiveDate to but what do I use to set the annotation meta:valDeprecationDate ?* For each template the use of meta:valEffectiveDate and meta:valDeprecationDate is mandatory; this covers, for the lowered templates in RDF, the timestamping requirements
Regards
Keith
Re: Hiding the (class_of)temporal part in template signature
Hans, I only can hope that your proposal will allow for concepts developed by IIP, JORD and my team.
-
- Posts: 283
- Joined: Sun Jan 22, 2012 10:02 pm
Re: Hiding the (class_of)temporal part in template signature
@Keith -
Ahh, that's it. Well, that unlocks a wider topic: when we load template instances in our store, what shall be done then? Important issue is the question how the system can find out whether the new information does or does not replace older information, and if so, which template then must be deprecated? But I will open another topic for that.
Ahh, that's it. Well, that unlocks a wider topic: when we load template instances in our store, what shall be done then? Important issue is the question how the system can find out whether the new information does or does not replace older information, and if so, which template then must be deprecated? But I will open another topic for that.