[ ] Dena: We can have various modelling decisions on it, depending on the application of the model. I assume the envisioned application is “ search and findability and reuse” therefore two approaches:
We can look at it as a controlled vocabulary, which can later be used as a filter in any search engine. I assume there are more vocabularies about roles that organizations can have in research.
a. It can be modelled as an instance of class (OrganizationRole).
I think both dataset and project can have funders.
a. class(Project Or Dataset) property(hasOrganization) class(Organization) .
b. class (Organization) property(role) class (OrganizationRole) > Class(Orgnization) has a property :role whose value can be :Funder.
c.The problem with this approach is traversing two hubs which can be seen as less performant.
Or it can be modelled as a direct relation.
a. Class(Project or Dataset) Property(Funder) class(Organization). > this can be seen better in term of performance.
In terms of URI, I am not sure if we, organizationally, are collaborating with the DCAT community and if we can request and extend their model using their namespace meaning if we are creating health-dcat: version? If not, then we should use the health-ri namespace.
Both approaches can be implemented with all the schema owl, rdfs, and shacl.
[ ] Jeroen: it is shown as mandatory for all metadata records; This in my opinion is wrong. I would have expected it to be part of a leave-metadataschema for funder and only if the leave is used it to be mandatory, together with the core-mandatory items. So I expect two metadata schemas: 1) the DCAT-AP core as schema for all metadata records and 2) a Funder-specific-leave-metadata-schema (with a table for funder specific items of which “funder” for now is the only mandatory item). A funder normally applies to funded research. But in principle no data is registered for free, the person doing it normally gets paid but do we see this a funding like in COVID by ZonMw or cancer research by KWF, or heart diseases by the Dutch Heart Foundation -> so a doctor in a hospital entering clinical data is funded by the hospital, should funder then be hospital Spaarne Hoofdorp??? I assume not so that is why I expect the leave for funder.
[ ] Anastasia: is it “funder” or “funding”? it is “funding” in the excel spreadsheet.
[ ] Dena: We can have various modelling decisions on it, depending on the application of the model. I assume the envisioned application is “ search and findability and reuse” therefore two approaches:
[ ] Jeroen: it is shown as mandatory for all metadata records; This in my opinion is wrong. I would have expected it to be part of a leave-metadataschema for funder and only if the leave is used it to be mandatory, together with the core-mandatory items. So I expect two metadata schemas: 1) the DCAT-AP core as schema for all metadata records and 2) a Funder-specific-leave-metadata-schema (with a table for funder specific items of which “funder” for now is the only mandatory item). A funder normally applies to funded research. But in principle no data is registered for free, the person doing it normally gets paid but do we see this a funding like in COVID by ZonMw or cancer research by KWF, or heart diseases by the Dutch Heart Foundation -> so a doctor in a hospital entering clinical data is funded by the hospital, should funder then be hospital Spaarne Hoofdorp??? I assume not so that is why I expect the leave for funder.
[ ] Anastasia: is it “funder” or “funding”? it is “funding” in the excel spreadsheet.