Open jonquet opened 5 months ago
@jonquet this is an interesting question. IMO, we should prepare nice documentation categorically specifying the MOD properties that apply to which objects (e.g., mod:SemanticArtifact, mod:SemanticArtifactCatalog, etc.). I am not sure whether we can solve this at the vocabulary level itself. Anyhow, I will work on this further, let me think for a while.
I originally used a very semantically loose property to encode the "mapping" between MOD properties. At that time (before MOD2 and adiption of DCAT) the properties where all supposed to apply on a unique object mod:Ontology.
Now mod:Ontology is discarded and since MOD2 we use mod:SemanticArtefact and mod:SemanticArtefactDistribution The dispatch of which MOD properties "applies" to mod:SemanticArtefact and whci ones applies to mod:SemanticArtefactDistribution was never really well done in RDF. Considering that some properties (as originaly defined without domain and range) may applies to both. See also #97
Now, with MOD3.2 we introduce metadata properties to describe another object (mod:SemanticArtefactCatalog) and then again some properties in MOD will apply to catalogue. Worst now, some "mappings" that were ok when we considered the property was only used to describe Semantic Artefact (or distrib) are not ok any more with catalogue... and even sometime we would like to distinguish the default property recommended by MOD to be used to describe a catalog. For example, to describe a SA the property schema:publishingPrinciples maybe used as a mapping of the main property mod:usedEngineeringMethodology .. but to describe a SAC we like to use the property schema:publishingPrinciples as main property to explain the principles.
@agbeltran @biswanathdutta @antony-wilson any idea how to represent this?