Issue 281: Erlagen OWL / CRM

Starting Date: 
2015-05-22
Working Group: 
3
Status: 
Open
Background: 

In 32nd crm sig meeting, the sig decided the existence of an official version of CRM in OWL. The Erlagen group will support this version.

Current Proposal: 

In 33rd meeting in Nuremeberg it is decided the following:

  • About URIs, Erlagen should communicate with FORTH how to find the other versions of Erlagen CRM/OWL.
  • The constraints proposed by Erlagen are accepted.
  • Only the owl official version will have two name spaces, CIDOC CRM and  Erlagen owl.
  • FORTH will produce the RDFS and ERLAGEN will produce the OWL. Both of them will be accessible through the CIDOC CRM site.
  • We do not create a new version if we have something that affects the instances.  
  • In the next meeting, we should make a decision about the name spaces.
  • In the CRM name space will be without restrictions.
  •  Erlagen will  present to the next meeting the properties that Erlagen have implemented as symmetric, in order to be approved by CRM-SIG.
  • The procedure is to be approved the transitivity and equivalence.
  • Manos Papadakis should check the transitivity for all Allen operators.  Christian Emil will review all CIDOC CRM transitive properties.

Nuremberg, May 22 2015

In 34th  meeting in Heraklio, the crm-sig reviewed the transitivity of properties. The outcome of the discussion is described here.

Also the sig decided that (a) P69: an example is needed, P150: an example in the scope note, P165: to add a note in the scope note about transitivity.

The discussion about symmetric properties is presented here. Symmetric properties are P69, P114,P121, P122, P130,P132, P133, P139. 

Heraklion, October 2015


Posted by Carlo  on 5/2/2016

In general, all properties are directed, in the sense that the objects connected by the property are connected in an ordered way, i.e. they form an ordered pair in which there is a first argument (drawn from the domain of the property) connected to a second argument (drawn for the range). This is similar to the notion of directed arc in a graph, in which case the nodes connected by the arc are distinguished as a source node and a target node. In this sense we can talk of the inverse of a property, which is another property whose connected objects are obtained by reversing each pair of objects connected by the property.

Would that work?
 


Posted by CEO on 18/2/2016
Dear Chryssoula (and Martin)
I am in the process of adjusting the scope notes for transitivity/intransitivity. The word 'transitive' is only used 10 times in the draft version you sent me and only in the amendment part. Should I adjust the scope notes in the main part an also add information  about intransitivity. This is a purely editorial question:
Should we write
This property is not transitive
Or
Should we leave the information out for intransitive properties.

I assume that if the property is transitive we add the following to the FOL part (and suppress the three assumed universal quantifiers as we to everywhere)
P(x,y)  & P(y,z)  -> P(x,z)

If the property is not transitive we drop the  FOL  information out since it would have required  existential quantifiers:
Exist x,y,z not(P(x,y) & P(y,z) -> P(x,z))

According to the homework I should also add a paragraph about transitivity in the introduction?



Posted by Chryssoula on 18/2/2016
On 18/2/2016 3:44 μμ, Christian-Emil Smith Ore wrote:

>Dear Chryssoula (and Martin)
>I am in the process of adjusting the scope notes for transitivity/intransitivity. The word 'transitive' is only used 10 times in the draft version you sent me and only in the amendment part.
>the draft version is the version we edited in the meeting, I have not updated yet!!. The decisions we took about transitive properties are in your notes that you have sent to me after the meeting.

>Should I adjust the scope notes in the main part an also add information  about intransitivity. This is a purely editorial question:
YES of course and I can make it, if the task is to add just a sentence in the scope note of the property, but (I am not sure), as far as I remember we have said that  if the transitivity or intransitivity is not obvious to the reader to add a comment under what aspect this  property is transitive or intransitive. I think these properties are with red letters in your notes!! but I repeat I am not sure!!


Posted by CEO on 18/2/2016
I will add info  about transitivity which is easy,  and also where intransitivity is not so obvious. Transitivity adds axioms to the theory and thus extend the deductive power. An axiom of intransitivity is an assumption of the existence of an instance. We may apply FOL CRM to a universe in which such an instance does not exist and still want the universe to be a valid model for the theory described by FOL CRM.  Therefore we should leave the intransitivity axiom out.


Posted by Chryssoula on 18/2/2016
Dear Christian Emil
I have added a sentence (This property is transitive) to the end of the scope notes of explicit transitive properties from your ppt. Below you may find my notes from the discussion. I am not sure, if I noted correctly about the P69, but I am sure that It is proposed that an example is needed. Also a comment is needed about the P165 (it is implicit transitive). Another note is about  P46, but I have not written why?


Posted by CEO on 19/2/2016
Dear Chryssoula,
I got caught by a cold yesterday and didn't manage to follow up. I  will do so during the weekend. The explicit transitive relations are unproblematic.
As I mentioned yesterday, Intransitive: Most of the intransitive properties are intransitive due to the simple technical  fact that the domain and range are different classes and most of them cannot be implicitly transitive.  For the possibly implicitly transitive that are definitely intransitive  we may add  a comment, but not add a FOL axiom since this will make the theory stronger than necessary.

The implicit or restricted ones need a short explanation. I will write the explanations and the paragraph for the term list by tomorrow evening.

In October the term recursive was introduced. In mathematics recursive differently and I think we should avoid the term. In general since we now have introduced FOL into the description we should stick to the use of terms in mathematical logic. More about that later.


Posted by CEO on 21/2/2106
Dear Chrysoula,
Please find a short note with  draft texts and suggestions attached.

I suggest that for all the explicitly transitive properties we add the sentence "The property is transitive" and the the fol formula
Pnn(x,y) & Pnn(y,z) ⊃ Pnn(x,z)

The unmarked properties should be considered to be intransitive.

When I read through the term list section, I noted that between some of the entries there should be added space (a blank line).  We may sit down in Prato and go through such editorial issues.

I will continue with the rest of my homework later. Now my daughter and I will go out skiing (down hill).


Posted by Christian Emil on 23/2/2016

I have worked with the problems Mark points out.

First of all, not symmetric is not equivalent to asymmetric see the quotes form wikipedia (could be any textbook):

In mathematics, a binary relation R on a set X is antisymmetric if there is no pair of distinct elements of X each of which is related by R to the other.

In mathematics and other areas, a binary relation R over a set X is symmetric if it holds for all a and b in X that if a is related to b then b is related to a.


P130 shows features of (features are also found on),
P139 has alternative form, P69 has association with (is associated with)

can all be not symmetric without being asymmetric and vice versa. My suggestion is to remove the symmetry  axiom form the FOL paragraph for the three properties and adjust the scopenotes,  This seems to be closest to the intention of scope notes and  best from  a mathematical/FOL point of view.  The way it stands now is close to what we criticise in  DC . WE cannot now the meaning without checking the value. The first order parts are not defintions of the properties, they state the stronges possible conditions that are true for all instances (or some if we use existential quantifiers). 


Posted by Martin 24/2/2016

On 23/2/2016 10:09 μμ, Christian-Emil Smith Ore wrote:
> I have worked with the problems Mark points out.
>
> First of all, not symmetric is not equivalent to asymmetric see the quotes form wikipedia (could be any textbook):
>
> In mathematics, a binary relation R on a set X is antisymmetric if there is no pair of distinct elements of X each of which is related by R to the other.
>
> In mathematics and other areas, a binary relation R over a set X is symmetric if it holds for all a and b in X that if a is related to b then b is related to a.
>
>
> P130 shows features of (features are also found on),
>   P139 has alternative form, P69 has association with (is associated with)
>
> can all be not symmetric without being asymmetric and vice versa. My suggestion is to remove the symmetry  axiom form the FOL paragraph for the three properties and adjust the scopenotes,  This seems to be closest to the intention of scope notes and  best from  a mathematical/FOL point of view.  The way it stands now is close to what we criticise in  DC . WE cannot now the meaning without checking the value. The first order parts are not defintions of the properties, they state the stronges possible conditions that are true for all instances (or some if we use existential quantifiers).
I agree:-)

In the 35th joined meeting of the CIDOC CRM SIG and   28th FRBR - CIDOC CRM Harmonization meeting the crm-sig proposed and decided the following:

- About the symmetric properties  P130 shows features of (features are also found on),  P139 has alternative form, and P69 has association with (is associated with) , it accepted the   proposal by CEO to  remove the symmetry  axiom form the FOL paragraph of the  properties and adjust the scope notes. 
- Also it is decided that for all properties that have same domain and range, and are not transitive, this should be stated explicitly.
- The crm-sig revised the scope note of P130.    
- In order to  avoid confusions and discussion about transitivity,   the scope notes of P165 incorporates (is incorporated in) and P150 defines typical parts of (defines typical wholes for) are revised appropriately. The changes are incorporated in CIDOC CRM 6.2.2
- The sig reviewed the text provided by CEO about transitivity and accepted to be added in the terminology section with the following additions: (1)  To put property numbers when reference is made to categories of properties i.e properties referred to time/space/transitive
(2)  To put class numbers when reference is made to them,. i.e In the examples referred to classes
- It is assigned to Christian Emil (1)  to write a definition   for "directed property"  (2) "symmetry" in order to be added into terminology. 

- It is still needed to be changed the symmetry statement in FOL

- It is still needed to review the statement about  “directed property” by Carlo (see Carlo's email on 5/2/2016)
- Symmetric properties are P69, P130, 139. The scope notes of P69 and P139 should be rewritten, by Martin, taking into account the statement of directedness.
- In the previous meeting has been accepted that in the official release we will keep things backwards compatible. (Compatible at least in instance level).   Mark should distribute to crm-sig the respective owl versions with compatible statements. (It is still open)

- About Erlagen version of OWL CRM it is accepted  the  headers proposed by Mark with the additional stipulation to Creative Common Licence. The headers proposed are described in this file. (see also issue 276)

 Prato, February 2016

 

posted by Christian Emil on 31/7/2016

Dear all
The condition 1,  the copyright of the ISO version, implies that we have to consider the draft versions, say version 6.2, as something so different from the ISO-version that the copyright of ISO does not apply. In case ISO sued us I think we will have a weak case if there is not an understanding by ISO. This includes the owl-versions since the scopenotes are included in the rdf. In the following I assume that ISO will not sue us.

The second condition implies that license should include commercial use and it should not be necessary for the licensee to share their adjustments, but they are allowed to do so. 


The corresponding Creative Commons license is CC BY 4.0


https://creativecommons.org/licenses/by/4.0/

    Share — copy and redistribute the material in any medium or format
    Adapt — remix, transform, and build upon the material
    for any purpose, even commercially.

    The licensor cannot revoke these freedoms as long as you follow the license terms.

Under the following terms:

·         Attribution — You must give appropriate credit, provide a link to the license, andindicate if changes were made. You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use.

    No additional restrictions — You may not apply legal terms or technological measures that legally restrict others from doing anything the license permits.


posted by Mark on 31/7/2016

Dear all,

the erlangen-crm currently is distributed under CC BY 3.0 https://creativecommons.org/licenses/by-sa/3.0/

Which means additionally they have to share the things they have built upon the erlangen-crm under the same license. In fact nobody up to now ever did that (not even attribution :( )and we did not sue anybody (like the british museum) ;-)
I think the main question is how we want to have it in future. Going to CC BY in fact is better for commercial things. However I think we should clarify what "appropriate credit" really looks like... And: Do we want to prosecute people who don't care about the license?


posted by Christian Emil on 31/7/2016

The core issue is "do we want to sue persons". The answer is no, we don't have the resources. The best we can hope for is that users of the standard give CRM-SIG "appropriate credit"​. A common way is write a credit template and publish it together with the license information.

 


posted by Mark on 31/7/2016

Ok - we should do that :)

Would you prepare something?

Mark


posted by Christian Emil on 31/7/2016

​I will try.

 

 

In the 36th joined meeting of the CIDOC CRM SIG and ISO/TC46/SC4/WG9 and the 29th FRBR - CIDOC CRM Harmonization meeting, the SIG continued the discussion about this issue and 

(a) asked   Steve to review the text prepared by GB and CEO for terminology section with the   additions indicated to the previous meeting.  

(b) decided that one of the chairs should write to ICOM and cc board mentioning the conversation that already happened approving this and asked CEO to talk with Monica in order to identify person in charge and then MD will send the  request to appropriate person.

 Iraklio, August 2016

Posted by Christian Emil on 29/3/2017

I sent the  email below to Martin in the beginning of December 2016 with the attached draft. We will need a status report from Martin from the meeting in Berlin in January.

--------

Dear Martin
Here are some updates on the license process