Description
A cleanup of relations describing parthood is necessary. The items to consider are as follows:
-- hasPart is asserted in DUL as Transitive (which is ok). In mereology, the parthood relation is also sometimes considered reflexive and antisymmetric (everything is a part of itself, but never a part of its proper parts). The reflexive/antisymmetric combo would be inconsistent in OWL-DL, nevertheless do we want to add either a reflexivity or antisymmetry axiom?
-- hasComponent is a subproperty of hasPart (which is ok) but it is not asserted as Transitive. Subproperties do not inherit this from superproperties, so if we want hasComponent to be Transitive (arguably, it should be, though I can see arguments for why not also) then this should be explicitly asserted.
-- hasEventPhase should be an Event-to-Event relation, and be a subproperty of hasPart, if it is to be used for parthood between Events.
-- hasResult is a subproperty of hasPart, asserted in EASE/SOMA.owl. It has no comments. Is this property used anywhere, can we (re)move it? A property to indicate posession of an outcome is useful, but should not be a parthood relation.
Most of the above are discussion items, and the decision could go in the end either way. In any case, I suggest we add comments to justify the choices in the OWL files themselves.