View unanswered posts | View active topics It is currently Sun Jan 19, 2020 4:32 am



This topic is locked, you cannot edit posts or make further replies.  [ 28 posts ]  Go to page Previous  1, 2, 3
 Class Model 

MMT SIG approved?
Poll ended at Tue May 29, 2012 7:52 am
Yes 20%   20%   [ 1 ]
Yes with comments 40%   40%   [ 2 ]
No 0%   0%   [ 0 ]
No with comments 20%   20%   [ 1 ]
Blank vote 20%   20%   [ 1 ]
Total votes : 5

 Class Model 
Author Message

Joined: Sun Jan 22, 2012 9:14 pm
Posts: 189
Hans uses the term "the RDL" for any ISO 15926 compliant triple store with web service endpoint. He is right to do so.

The term used for "THE RDL" in part 8 is the Core Library or Core RDL.

Indeed the Core RDL only has core classes and reference individuals (but much more that the part 4 spreadsheet content)

A "federation" is a logical database formed of cooperating endpoints on the internet, but shielded from users outside the federation's group. These endpoints only contain self-owned data (aside from their data cache) and use the other endpoints to refer to.


Mon May 21, 2012 8:05 am
Profile

Joined: Thu May 17, 2012 10:50 am
Posts: 4
Onno,
I am not sure we should be confusing the phrase of a RDL. A Reference Data Library should refer to Reference Data rather than to ANY compliant triple store. That may be OK for MMT users, but for general viewers they need to recognise the difference between Reference Data which must be distributed and available by its nature and project data, which may not need to be distributed. I know there are subtle differences and indeed as we start to deploy there will be different requirements, which I think Hans has attempted to address in recognising that the lifecycle of a class is different from the lifecycle of instance data.

So back to the thread.

Is there not a method of doing this by CLASSIFICATION ? From what I read, we need to visit every class in the RDL whether Core or otherwise and create UR super classes.

Also within the RDL, if a user was to make a modification, there is a positive action required, and they will generally know that a CLASS has been modified or superceded and therefore the CLASS can either be reclassified as INVALID and the new class (I assume this will be a specialism or the class has been retired) marked as VALID. As you correctly state, how will the instances that use the classification know the class is now invalid ? We have to postively look for the UR class, so why not a CLASSIFICATION class. If there is no VALID class, then the class is not VALID ?

In either case they will require time stamps.

I am not sure what the addition of an UR super class brings over a VERSION or VALID classification.

In all cases we need to ensure that the class is relevant.

In all cases we are adding to a processing overhead.


Fri May 25, 2012 6:51 am
Profile

Joined: Thu Mar 08, 2012 12:39 pm
Posts: 16
Dear MMT member,

To contemplate the presented model, as Victor Agroskin told(Message May 16) the template for the class CompositionOfOIM will have a template statement as a role in a template and It is not complience with part 7.
I was thinking that the Templates Statements would be conected by ClassOfCompositionOfIndividual. In this way the OIM parts would be implicit, and it will only generate more difficulty to write sparql queries than in the model presented.

Best Regards
Geiza


Fri May 25, 2012 3:54 pm
Profile

Joined: Sun Jan 22, 2012 10:02 pm
Posts: 283
Dear Geiza,

If you have an instance of ClassOfInformationRepresentation in the role of classOfWhole in an instance of ClassOfCompositionOfIndividual, then the classOfPart role better refers to an instance of ClassOfInformationRepresentation (or a subtype thereof) as well. Otherwise you may get nonsense.

This weekend I will, as promised to Victor, post a document in which I propose a slight correction of Part 8, in that a TemplateStatement not only is a subclass of MultidimensionalObject and ClassOfInformationRepresentation in the text of Part 8, but also in its OWL code.

The very moment we have consensus about this we can do what you proposed.

Regards,
Hans


Sat May 26, 2012 4:06 pm
Profile

Joined: Sun Jan 22, 2012 10:02 pm
Posts: 283
Through the regular e-mail I received this question from Pavel:
Quote:
1) Could you give some examples of usage of the Class Model Template? 2) As I understand you want to use it to store historical data not only for process plants but also for RDL objects. So if we want to say if our class is deprecated - we have to create a subclass which is classified by status DEPRECATED on 10/10/10? And there would be also subclasses for each lifcycle stage of this class


My response to him was:
Quote:
Examples can be found at
http://www.infowebml.ws/mapping/mapping-linelist.htm and the bottom part of the OWL file derived from that: http://www.infowebml.ws/mapping/line-list-data-for-line-RZ17801.owl

The idea is this: if we have a class, be it high in the hierarchy (RDL) or P-101 in our plant design, changes happen to the definition of that class and those changes make it a different class. But that is not manageable for the user. If he refers to the class PUMP he refers to its URI, and if he refers to P-101 he also refers to that URI.

That is why I have proposed the idea of an Ur-Class that never changes, but that has subclasses that are involved in the information represented by a template. That U-Class is what you always refer to, and your software should collect the latest status of that class by collecting templates in which that class plays an Ur-Class role. Thereby that software shall leave older templates, of which the information content has been changed, out of that collection.

That collection is attributed to yet another subclass of the applicable Ur-Class. See also http://www.15926.info/plant-design/index.htm

In case someone would be interested in the lifecycle of the class definition, all the older information is available, including the period in time that a particular template was valid.


Sat May 26, 2012 4:15 pm
Profile

Joined: Thu Mar 08, 2012 8:37 am
Posts: 13
Location: Haarlem, the Netherlands
It is indeed quite confusing due to the paricular example that Hans used right in the beginning of the thread. The approach looks OK to the depth of my knowledge (though several "issues" were noticed by Victor). But that approach for dealing with class "maintenance" during their lifecycle is mixed here with whether designing of a plant should be done at individual or class level which is arguable.


Mon May 28, 2012 12:23 pm
Profile

Joined: Tue May 15, 2012 8:48 am
Posts: 77
Mikhail Fedorov wrote

"It is indeed quite confusing due to the paricular example that Hans
used right in the beginning of the thread. The approach looks OK to
the depth of my knowledge (though several "issues" were noticed by Victor).
But that approach for dealing with class "maintenance" during their
lifecycle is mixed here with whether designing of a plant
should be done at individual or class level which is arguable."


This is also my problem

If the example chosen had been one where a specialized class
is clearly appropriate such as particular pump model in a
manufacturers catalogue I would have no problem approving
this proposal.

Keith


Tue May 29, 2012 10:29 am
Profile
Site Admin

Joined: Wed Jan 18, 2012 8:50 pm
Posts: 32
LOCKED this topic.

A new topic has been started.

See viewtopic.php?f=3&t=19


Mon Jun 18, 2012 11:50 am
Profile
Display posts from previous:  Sort by  
This topic is locked, you cannot edit posts or make further replies.   [ 28 posts ]  Go to page Previous  1, 2, 3

Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
cron
eXTReMe Tracker
Powered by phpBB® Forum Software © phpBB Group
Designed by ST Software.