Page 2 of 3

Re: p7tpl file generation

Posted: Mon Nov 12, 2012 9:43 pm
by vvagr
And one other important issue - valFOLCode should be an OWL representation of some concept defined by Part 6. Which probable should have some alternative representations - in Part 2 language and as a template.

Re: p7tpl file generation

Posted: Tue Nov 13, 2012 9:43 am
by HansTeijgeler
I have on idea what you mean with that.
Please explain

Re: p7tpl file generation

Posted: Tue Nov 13, 2012 6:39 pm
by vvagr
Hans,

Part 6 defines various types of data "about" reference data items. Definitions, identifications, administration data. This data is first defined in terms of Part 2 relationships, and also as a template for that relationship. Some relationships can be further mapped to RDF/OWL as properties - subClassOf, type. For example, there are some useful ways to express in RDF language identifications - @en, @ru, etc.

So it should work other way around - if there is p7tm:valFOLCode property, there should be a way to express the same relationship as template TemplateFOLCode, for example, and also as an instance of ClassOfDefinition, itself classified by reference data class FOL AXIOMS or something like that. Other then RDF/OWL representations of ISO 15926 data can not be deprived of FOL representation ways.

Re: p7tpl file generation

Posted: Tue Nov 13, 2012 8:26 pm
by HansTeijgeler
As far as I know Part 6 isn't approved yet.
May I invite you to make a detailed proposal for the way it should be done?

Re: p7tpl file generation

Posted: Tue Nov 13, 2012 8:44 pm
by vvagr
Yes, and it is evolving rather fast, although I've difficulties getting access to drafts. The team behind it has their own ideas, as far as I understand, and Onno is in a better contact with them. Probable they can together think of a good way to do that.

Personally I think that suggested Prover 9 notation code in an RDF property is only a partial solution to the problem. The reason is simple - the template iitself is identified by its URI, but all templates in the axiom are identified only by name.

Of course it is not a problem if only base and prototemplates from a standard set are included in the axiom.

But imagine that I'm defining some company-level template CompositionOfThisAndThat with URI http://xyzco.com/templates#CompositionOfThisAndThat and I need to use my template CompositionOfThis in FOL axiom. How can I use its URI to show unambiguously that it is indeed http://xyzco.com/templates#CompositionOfThis?

I don't know good answer for that on ISO level, but I'm doing some implementation for it in my Editor. Just like there are no standard JORD pattern descriptions, but there is a big IIP pattern database implemented in software.

Re: p7tpl file generation

Posted: Tue Nov 13, 2012 8:56 pm
by OnnoPaap
Yes part 6 is sent up for publication to the ISO secretariat, I heard today in the web meeting of the T25 meeting (click meetings)

No problem. Our group is there to generate issues and to solve them.

So Victor it would be great if you would make a model out of your solution for FOL, that we can discuss and change the p7tpl accordingly. We will implement it and place an issue on the part 6 forum.

By the way, this may seem like a part 8 or even part 9 scope. We can go that direction as well.

Re: p7tpl file generation

Posted: Wed Nov 14, 2012 9:31 am
by HansTeijgeler
Victor,

I propose to use the template DefinitionOfClassWithExternalCode (see http://15926.org/templatespecs/CL-DEFN-01a.xml).

The application of a specialization of this template can be found at http://www.infowebml.ws/Part3/addendum2.htm

Re: p7tpl file generation

Posted: Mon Nov 19, 2012 2:54 pm
by OnnoPaap
In summary, after Victor's observations the following:

Templates proposed (to the MMT SIG) on the 15926.org have the namespace

Code: Select all

proposedtpl="http://15926.org/templates-test/templates#"
Templates approved then go to namespace

Code: Select all

p7tpl="http://standards.iso.org/iso/ts/15926/-8/ed-1/tech/reference-data/templates#"
We will stick to namespaces given out by ISO 15926-8 namely:

Code: Select all

dm="http://standards.iso.org/iso/ts/15926/-8/ed-1/tech/reference-data/data-model#"
p7tpl="http://standards.iso.org/iso/ts/15926/-8/ed-1/tech/reference-data/templates#"
p7tm="http://standards.iso.org/iso/ts/15926/-8/ed-1/tech/reference-data/template-model#"
Obviously pending the changes we vote for, about shorter namespace addresses.

If agreed, then I will update the program that consolidates the templates to an OWL file.

Edit: for valFOLCode we need to model a solution

Re: p7tpl file generation

Posted: Mon Nov 19, 2012 3:02 pm
by vvagr
Thank you.

I support an idea of valFOLCode, the one change I'm suggesting is that it is moved to p7tm: namespace.

Re: p7tpl file generation

Posted: Mon Nov 19, 2012 3:32 pm
by HansTeijgeler
OK with me.