2

I want to say that in my ontology each individual has exactly one most-specific type.

In this case:

  1. Is it accurate to say my ontology does not allow multiple inheritance? I don't quite understand what multiple inheritance means in ontologies. Is this refer to rdf:type or rdfs:subclassOf? because I don't mind if a concept has multiple parents but I don't want individuals to have multiple types.

  2. Is it accurate to say in my ontology all classes are pairwise disjoint, or should I say for instance in the ontology classes are pairwise disjoint unless they are participating in subsumption relation with each other?

  3. Is there a better terminology or way to actually describe such ontologies?

Bahar
  • 770
  • 5
  • 18
  • This document uses the word "inheritance": https://www.w3.org/TR/sw-oosd-primer/ – Stanislav Kralin Oct 29 '19 at 12:33
  • 1
    Can you restate the question without using the term "participates in". This doesn't have a defined meaning in OWL and I don't think you need it. You could say each individual has one most-specific type, all other types will be inferred via subClassOf axioms. – Chris Mungall Oct 29 '19 at 23:58
  • Thanks that make sense. I have updated it – Bahar Oct 30 '19 at 13:30

1 Answers1

0

In literature you can find the expression "multiple inheritance" referring to both the extensional (rdf:type) and intensional (rdfs:subclassOf) meanings. RDF/OWL natively support both of them.

  1. No, in general this is not correct. I don't think there is an official (and unambiguous) term for "individual being instance of multiple classes".
  2. Your first statement is a sufficient condition, while the second one is necessary and sufficient. Alternatively, you can say that if two classes share an individual, then one of the two is a subclass of the other one.
  3. I think that "one most-specific type" is already a well-describing terminology.
logi-kal
  • 7,107
  • 6
  • 31
  • 43