Issue 590: LRMoo: review of properties--request for comments

ID: 
590
Starting Date: 
2022-04-28
Working Group: 
3
Status: 
Done
Background: 

The LRMoo WG has done a review of technical aspects of the LRMoo property definitions. Namely, quantifiers, as well as transitivity, symmetry, reflexivity as discussed in recent SIGs for CRMbase.  

We have found some cases of questionable quantifiers.

Finally, we reviewed the handling of the .1 properties, based on the best practices recently discussed on the list. 

 

At this point we'd appreciate anyone who would have time to review our conclusions.

These are found in a 3 page document here:

https://docs.google.com/document/d/1u7bjA8ttGHGXRQ04PaYSGzsicyMpzc_PD_aQj5cDyx4/edit?usp=sharing

 

This request for comments is open until May 14 (that is until after the upcoming SIG meeting).

 

Many thanks, Pat

Current Proposal: 

In the 53rd CIDOC CRM & 46th FRBRoo SIG meeting, the SIG the Sig went through the revised the table of quantifications, fol axioms, (.1) properties, and declaration of new sub/superproperties for the set of LRMoo properties (HW by PR & MZ) and edited some of them. The updated list can be found here. 

 

Decisions:

  • label for the quantification (1,n:1,n): many to many, necessary, dependent. To be added to the list of possible quantifications in LRMoo.  
  • Re. transitivity, symmetry and reflexivity: the non-symmetric/reflexive attributes mean to say that some instances of the property can be symmetric/reflexive, whereas some others can be asymmetric/irreflexive. For properties with the same domain and range, a statement must be offered wrt. their transitivity (+, -, neither).
    However:
    If a property is known to be reflexive/symmetric for all its instances, then it should be marked as such. The same applies to properties known to be irreflexive/asymmetric for all their instances.

Conversely, if a property is neither symmetric/reflexive nor asymmetric/irreflexive, it must not be given a characterization regarding these attributes. Non-x statements should be avoided.

  • Re the .1 properties: deprecate the R33.1 has encoding property, it can inherit the one from P3.1 has note.
  • Re. declaring new sub/super properties:
    • R2 is derivative of IsA R68 is inspired by, on the grounds of it being much more specific. 
    • R2 is derivative of not declared a subproperty of P130 shows features of, as R68 (its direct superproperty) is listed as a subproperty of P130.

 

May 2022

Outcome: 

In the joint 54th CIDOC CRM & 47th FRBR/LRMoo SIG meeting, Pat Riva informed the SIG that the issue has closed by the editors' decsion, on the grounds of there not being any pending remarks objecting the proposed quantifications of LRMoo properties. 

 

Rome, September 2022

Reference to Issues:

360
LRMoo