Implicit subclass links due to the subproperty domain/range inheritance

[mereolog]

There are three implict subclass relationships due to the inheritance of the domains for subproperties.

In the table below:

• property1 rdfs:subPropertyOf ?property2
• property1 rdfs:domain ?domain1.
• property2 rdfs:domain ?domain2.
• but not ?domain1 rdfs:subClassOf ?domain2.

property1	property2	domain1	domain2
idmp-mprd:isAuthorizedThrough	cmns-pts:undergoes	idmp-mprd:AuthorizedParty	cmns-pts:Undergoer
cmns-org:isMembershipPartyIn	cmns-pts:actsIn	cmns-org:MemberBearingOrganization	cmns-pts:Actor
cmns-org:isOrganizationMember	cmns-pts:undergoes	cmns-org:OrganizationMember	cmns-pts:Undergoer

These facts should be reviewed as we shouldn not keep such subclass links implict: so either add them explicitly, i.e., as triples, or change the property infrastructure.

Note that there is a similar situation with respect to the range inheritance - see the table below:

property1	property2	range1	range2
idmp-mprd:hasAuthorizedParty	cmns-pts:hasUndergoer	idmp-mprd:AuthorizedParty	cmns-pts:Undergoer
cmns-org:hasMembership	cmns-pts:hasActor	cmns-org:MemberBearingOrganization	cmns-pts:Actor
cmns-org:hasOrganizationMember	cmns-pts:hasUndergoer	cmns-org:OrganizationMember	cmns-pts:Undergoer

[ElisaKendall]

@mereolog These are all related to property chains, so we would need to make sure that any additions would not violate restrictions in OWL related to complex properties.

[ElisaKendall]

@mereolog The other thought I've had is that there may be cases where someone that is an authorized party in some situation then delegates that authorization to another party, and in that case they become the actor, the authorizing party, not the undergoer, but I need to test that to see if it really works to chain those together, and what, if any, impact that would have on the model.