This focus, based on a very high level of expertise and analytical acumen, brings great clarity and operationality.
But if the structural dimension is reserved for ontology (and classes/instances...), how can we relate concepts, independently of the rather poor properties (labels..., ["what things are “like” ?]) and relationships (broader, related..., [formal subsumption ?]) offered by SKOS ?
For example, "AI", "linguistics," "computer science," "neuroscience"...?
Moreover, can't there be any redundancy if one introduces rdfs:subclass ?
cbCharacter:concept and mvCharacter:concept in the above example are within the Comic Book Character and Movie Character ontologies respectively, but they point to the same conceptual character Batman.
Note that there is another usage of the term Batman, as a soldier asssigned as a valet or orderly in the British Army (it's a usage that's becoming obscure).
concept:Batman_Valet a skos:Concept ;
skos:prefLabel "Batman"@en ;
skos:altLabel "Valet"@en, "Manservant"@en ;
skos:description "A manservant or valet to an officer in the British Army in the 19th and early 20th centuries."@en ;
.
While it's stretching things a bit, you could see Alfred in this role:
comicBookCharacter:AlfredPennyworth a class:ComicBookCharacter ;
This focus, based on a very high level of expertise and analytical acumen, brings great clarity and operationality.
But if the structural dimension is reserved for ontology (and classes/instances...), how can we relate concepts, independently of the rather poor properties (labels..., ["what things are “like” ?]) and relationships (broader, related..., [formal subsumption ?]) offered by SKOS ?
For example, "AI", "linguistics," "computer science," "neuroscience"...?
Moreover, can't there be any redundancy if one introduces rdfs:subclass ?
This is where you explicitly define within an ontology a linkage to a concept:
Shape:ComicBookCharacter a sh:NodeShape ;
sh:targetClass class:ComicBookCharacter ;
sh:property [
sh:path cbCharacter:concept ;
sh:class skos:Concept ;
sh:nodeKind sh:IRI;
];
.
Shape:MovieCharacter a sh:NodeShape ;
sh:targetClass class:MovieCharacter ;
sh:property [
sh:path mvCharacter:concept ;
sh:class skos:Concept ;
sh:nodeKind sh:IRI;
];
.
comicBookCharacter:Batman a class:ComicBookCharacter ;
cbCharacter:concept concept:Batman ;
rdfs:label "Batman";
.
mvCharacter:Batman a class:MovieCharacter ;
mvCharacter:concept concept:Batman ;
rdfs:label "Batman";
.
concept:Batman a skos:Concept ;
skos:prefLabel "Batman"@en ;
skos:altLabel "The Dark Knight"@en, "Bruce Wayne"@en ;
skos:description "A guy who likes cosplaying a giant bat and beating people up at night."@en ;
skos:inConceptScheme conceptScheme:FictionalCharacters ;
.
```
cbCharacter:concept and mvCharacter:concept in the above example are within the Comic Book Character and Movie Character ontologies respectively, but they point to the same conceptual character Batman.
Note that there is another usage of the term Batman, as a soldier asssigned as a valet or orderly in the British Army (it's a usage that's becoming obscure).
concept:Batman_Valet a skos:Concept ;
skos:prefLabel "Batman"@en ;
skos:altLabel "Valet"@en, "Manservant"@en ;
skos:description "A manservant or valet to an officer in the British Army in the 19th and early 20th centuries."@en ;
.
While it's stretching things a bit, you could see Alfred in this role:
comicBookCharacter:AlfredPennyworth a class:ComicBookCharacter ;
cbCharacter:concept concept:Batman_Valet, concept:AlfredPennyworth ;
rdfs:label "Alfred Pennyworth";
.
As to subclass inheritance, concepts are not subclassed, they are usually part of concept schemes.
Striking and valuable discourse - thanks.