The relationships between terms in even a small vocabulary (3 or 4) may involve
an is_a or part_of relationship which might need to be explicit in a query.
Managing this by relationships between terms appears to be the right way to go.
Suggest adding a term relationships section to the ontology.
If an archetype is "a" fundamental concept in clinical terms then I do not (easily) see why you will need to establish relationships in the ontology. Can you provide one or more examples to open my eyes?
I was a strong proponent of this some time ago but now I'm increasingly convinced we should leave this to "external ontologies". The token of "ontologies of reality" vs. "ontologies of information capture" applies. That said I think there is a need to depict relationships that relate to the latter aspect: e.g. how a data item is related to another for the purpose of information representation, one practical example might be to disable smoking quantity item if the patient is a non-smoker. Some could argue this is a presentation issue but I think this is a semantic issue and has to be dealt with at this level.
information representation, one >>practical example might be to disable
smoking quantity item if the patient is a non-smoker.
Many such rules will cross over archetypes. I wonder if GDL might have a
We need to improve the assertion definition in archetypes, probably some parts of GDL could be reused for this.