Open sdevalk opened 2 years ago
One would argue that 1. and 2. are the same relationship, just looking at it from a different direction. You would not want to search based on two properties, so better to stick to one property for both. The same applies to 3. and 4. And also to 5. and 6.
Having said that, the CRM allows modelling such relationships through events and activities. E.g. you need a record of an activity for the education of artist A and then link artist A and artist B to this activity as contributors in different roles.
Another example: the activity of painting of artist A was influenced by the activity of painting (or an artwork produced) by artist B.
Does this sound reasonable?
@natuk Thank you - much appreciated!
Yes, that sounds reasonable!
If I translate this to RDF/Turtle, it could look like:
<https://some-education-activity>
a crm:E7_Activity ;
crm:P2_has_type <http://vocab.getty.edu/aat/300191862> ; # "Educating"
rdfs:label "Artist A educated Artist B" ;
crm:P9_consists_of [
a crm:E7_Activity ;
crm:P2_has_type <http://vocab.getty.edu/aat/300069743> ; # "Teaching"
crm:P14_carried_out_by <https://artist-a> ;
] ;
crm:P9_consists_of [
a crm:E7_Activity ;
crm:P2_has_type <http://vocab.getty.edu/aat/300449145> ; # "Learning"
crm:P14_carried_out_by <https://artist-b> ;
] .
This is (my understanding of) the 'role pattern' (see e.g. https://github.com/linked-art/linked.art/issues/493).
(Alternatively, a construct with crm:PC14_carried_out_by
and crm:P14.1_in_the_role_of
could be used, but that's not used/recommended by Linked Art.)
An other approach, perhaps, would be to use an attribute assignment that I've found in this PR by @azaroth42. In RDF it could look like:
<https://artist-a>
a crm:E21_Person ;
rdfs:label "Artist A" ;
crm:P140i_was_attributed_by [
a crm:E13_Attribute_Assignment ;
rdfs:label "Assign Artist B as Student of Artist A" ;
crm:P141_assigned <https://artist-b> ;
crm:P177_assigned_property_of_type <http://vocab.getty.edu/aat/300025909> ; # "Student"
] .
--> vocabulary terms for the different relationships in common use
I think this becomes part of #186 ?
Hi all,
I'm modeling a dataset with biographical information of artists and using the Actor patterns as foundation.
My dataset also contains relationships between actors. For example:
I realize Linked Art does not aim to capture all possible relationships, and that this topic has been discussed before. But I was wondering: do you have a recommendation for modeling this information (e.g. using Linked Art, 'pure' CIDOC-CRM, or perhaps another ontology)?
Thank you.