Doubts in development of a basic ontology using RDF/S: ontology reuse and "instance vs. inheritance" -

i trying develop simple "ontology" using rdf , rdf schema.

for now, have written following classes , properties:

<rdfs:class rdf:about="#model">...</rdfs:class> <rdfs:class rdf:about="#system">...</rdfs:class> <rdfs:class rdf:about="#concept">...</rdfs:class> <rdfs:property rdf:about="#represents">     <rdfs:domain rdf:resource="#model" />     <rdfs:range rdf:resource="#system" /> </rdfs:property> <rdfs:property rdf:about="#includes">     <rdfs:domain rdf:resource="#model" />     <rdfs:range rdf:resource="#concept" /> </rdfs:property> 

the idea i'd define "model" (i.e., model instance) set of "concept"s. thus, such "model" instance "represent" "system": "system" can described using "concept"s defined in "model" (i may create class "systemdescription" purpose).

now, doubts following:

1. i'd define concepts subclasses of "concept", "includes" property need them instances in order put in relationship "model" instance; in other words, concepts have both rdf:type , rdfs:subclassof set "concept" -- approach? (it seems me problem requires concepts both instances , classes)

2. suppose i'd reuse skos (simple knowledge organization system) ontology: ok define "model" subclass of skos:conceptscheme (or "concept" subclass of skos:concept)?

and, less specifically, have advice approach considering such modeling problem?

your questions, in order:

1) technically possible of course, in opinion, not approach. make ontology more complex, both human-understanding perspective logical perspective: puts ontology in complexity class (owl full) reasoning tools not deal with.

there mechanism in owl 2 (notably 'punning' mechanism) provide workaround problem in specific cases. however, add complexity, don't understand why want this. have class concept, whole point of being class has individual concepts instances. why want them (also) subclasses?

2) yes, absolutely fine.

as general advice: think approach have looks fine. kiss principle applies ontology modeling as design excercise. lot depends on actual requirements have - see earlier remark concepts becoming subclasses: if have reason want that, may need alternative approach, without knowing reasons requirement it's not possible recommend alternative.