Open JDMKSZ opened 1 month ago
Some remarks:
The term "legal person" can be ambiguous because it is often used as a synonym of terms that refer only to non-human legal entities, specifically in contradistinction to "natural person".
An entity that is able to carry out actions. Typically either a natural person or an organization.
; so not limited to legal persons in this definition.Ok so in fact we already have Agent as "superclass". During our latest meeting, I noted to create a request for "Legal person", but I couldn't remember why exactly. I think it was because due to the new concept of NihdiNumber, which identifies either a person (health care worker) or an organization (hospital, ...). So it might be an identifier for (a subset of) Actor then?
I think we should ensure alignment in definitions and (formal) semantics with the SEMIC Core vocs (https://github.com/SEMICeu/) which are the basis for the ICEG and OSLO data models.
Such semantical issues should be discussed over there in the first place. That is why they are created and maintained. Creating properties independently from those creates unnecessary mapping challenges.
my 50cents,
Bert
I think we're largely aligned with the international standards currently in fedvoc, but we may need to realign with latest SEMIC Core Business 2.2.0:
For Belgian-specifics: maybe drop the draft CbeRegisteredEntity class ; it's a specific variant of "FormalOrganization" for CBE (not LegalEntity as it also contains some non-legal organizations)
A unifier of person+organization would indeed be foaf::Agent. Not sure if more specific subclasses are needed e.g. for:
there is no need for things like CbeRegisteredEntity (or "Belgian" whatever)...
Not sure if more specific subclasses are needed e.g. for:
* "legal person" = person + legal entity * "health care provider" = a person or organization that provides health care services
That is the purpose of the semantical models: if there is a need (use-case) in which the scope is restricted to heatlth care providers (those that are registered) and these have many additional properties than "normal" agents, then it can be considered to introduce them. But if, for instance, the only requirement would be that a health care provider is to be registered, then the mandatoriness of the registration is sufficient to make a distinction. And then the term is not needed.
In that case the modeling advice for health care could be that when they introduce in their data models they should apply this pattern of contraints. This is the discussion between a Vocabulary (a list of reuseable terms) and an Application Profile (the usage of terms in an application context). Mixing both in one document leads always to a form of confusion.
As I do not know if FedVoc is a Vocabulary or an Application Profile, and the intented use of the introduced terms, it is hard to give a firm advice.
Following these criteria:
I'd say FedVoc is a Vocabulary (containing different ontologies) rather than an application profile.
Linked to "organization" and "person", I think there's a concept that could be added to the vocabulary, and that is that of a "Person" as "superclass" of a natural Person, and a so-called Legal person" ("rechtspersoon" / "personnalité juridique"). In several exchanges, a.o. FourthWay, an actor can be either a natural person or a legal person.