Is Entity an instance of SetOrClass?

[arademaker]

Whenever I write (instance X Entity) it implies, by the definition of domain and the axiom (domain instance 2 SetOrClass) that (instance Entity SetOrClass).

One can also writes (instance X SetOrClass) and this implies the axiom (instance SetOrClass SetOrClass) by the same reason. Double-checking if that makes sense.

[arademaker]

I believe that [1,2] help us to understand that this type system of SUMO works just fine. Copying here for references:

[1] I. Horrocks, “Three Theses of Representation in the Semantic Web,” pp. 1–9, May 2003.
[2] P. Hayes and C. Menzel, “A Semantics for the Knowledge Interchange Format,” pp. 1–9, May 2001.

[arademaker]

BiologicalSpecies is the only symbol subclass of Class, is it right?

From (1) partition SetOrClass Set Class and the domain of partition we have

(2) instance Set Class
(3) instance Class Class

From (4) subclass Set SetOrClass and the domain of subclass we have

(5) instance Set SetOrClass
(6) instance SetOrClass SetOrClass

Note that the axiom (subclass Class SetOrClass) was commented to avoid the instance SetOrClass SetOrClass that we already have from (6) above. If this axiom was uncommented, we would also have

(7?) instance Class SetOrClass
(8?) instance SetOrClass SetOrClass

But we also have the rule

(9) instance ?c class <=> subclass ?c Entity

So from (1) and (2) we also have

(10) subclass Set Entity
(11) subclass Class Entity

and from (10) and (11) and the domain of subclass:

(12) instance Set SetOrClass
(13) instance Class SetOrClass

considering the domain declaration for the predicate instance, we also have from the above axioms:

(14) instance Set Entity
(15) instance Class SetOrClass
(16) instance Class Entity
(17) instance SetOrClass SetOrClass

Possible problems:

1. I don't see why we need to remove the axiom (subclass Class SetOrClass).
2. Set is an instance of Entity but also a subclass of Entity.
3. Class is an instance of Entity but also a subclass of Entity

[arademaker]

and I can see how to derive subclass Class SetOrClass

[arademaker]

Related to it:

http://sigma.ontologyportal.org:8080/sigma/Graph.jsp?kb=SUMO&term=Class&inst=inst&lang=PolishLanguage

Class is a 'concept' isolated in SUMO, it is not a subclass of SetOrClass and it contains one single subclass: BiologicalSpecies. Does it make sense? Why BiologicalSpecies is so different to be disconnected from the other classes?

[arademaker]

Commit a8f343f is one step for this issue.