Issue 560: Automatic generation rules for deriving the RDF form from the CIDOC CRM definition document
In the 50th joint meeting of the CIDOC CRM SIG and SO/TC46/SC4/WG9; 43nd FRBR – CIDOC CRM Harmonization meeting, Elias Tzortzakakis presented the steps he followed to produce the RDF for CIDOC CRM v7.1.1 and identified a number of issues that need tending to.
The current issue concerns the rules for deriving the RDF from the CIDOC CRM definition document in an automated manner.
Its main focus is to provide answers to the following questions:
- Should the official RDF still contain P81 and P82 pointing to literal or should these be substituted by P81a/b and P82a/b that point to literals? It can yield alternative treatments of time intervals.
- Should this policy be also implemented for P170i? Elias maintains that this a/b split of the property should be implemented for P170i as well.
- For the moment we have a separate module for PCs –should they be integrated in the official RDF?
- E61 Time Primitive should it be xsd:dateTime in the official RDF?
- Datatypes of Primitives that are specified as either xsd datatypes or literals. However, a decision must be reached re. E94 Space Primitive and E95 Spacetime Primitive. He needs feedback from the SIG for that.
June 2021
In the 51st CIDOC CRM & 44th FRBRoo SIG meeting, SIG members decided to formally close the issue on the grounds of all its sub-topics having been resolved (with the exception of the point about the module containg the PC properties, which will be made part of a new issue in due course).
Details on the points addressed below:
In the 51st CIDOC CRM & 44th FRBRoo SIG meeting,
In the 51st CIDOC CRM & 44th FRBRoo SIG meeting, upon discussing issue 555 and specifically working towards automatically generating a JSON-LD context, the SIG decided to start a new issue where to determine other serializations that it might useful to autogenerate for different audiences.
- Should the official RDF still contain ‘P81 ongoing throughout’ and ‘P82 at some time within’ pointing to literal or should these be substituted by P81a/b and P82a/b that point to literals? It can yield alternative treatments of time intervals.
⇒ Just keep P81 and P82 pointing to literal and also include P81a/b and P82a/b
- Should this policy be also implemented for ‘P170i time is defined by’? Elias wonders if the a/b split of the property should be implemented for P170i as well.
⇒ We only have P170i and do not include a/b properties
- For the moment we have a separate module for PCs –should they be integrated in the official RDF?
⇒ There are separate modules for version 6.0 and 6.2.
⇒ We need one for 7.1.1 (which includes a new property of property: P189 approximates / P189.1 has type)
⇒ FORTH plans to open a new issue on this
- E61 Time Primitive should it be xsd:dateTime in the official RDF?
⇒ Just keep literal and avoid restriction to specific datatypes
- Datatypes of Primitives that are specified as either xsd datatypes or literals. However, a decision must be reached re. E94 Space Primitive and E95 Spacetime Primitive. He needs feedback from the SIG for that.
⇒ Just keep literal and avoid restriction to specific datatypes
Issue closed