Clarify definitions of navigational parent and child predicates

[rjyounes]

gist 9.7.0 defines the following predicates (the latter two are for non-polyhierarchical trees):

hasSubCategory
hasNavigationalChild

hasSuperCategory
hasNavigationalParent

hasUniqueSuperCategory
hasUniqueNavigationalParent

I propose removing the navigation predicates, as I don't see any use for them that can't be handled with the category predicates. Judging from the names, they may have been created for faceting, but they still seem redundant to me.

[dylan-sa]

In the 12/16 issue review meeting, we decided to keep the "navigational" predicates and tweak their skos:definitions to better indicate purpose. These predicates can be useful in some contexts (e.g., within a user interface) when the child category is not necessarily an instance of the parent category but it is nevertheless helpful to represent the relationship hierarchically. (For example, refrigerator handles aren't refrigerators, but it could be useful to group them under the refrigerator category.) Following up with a PR for this one shortly.

[uscholdm]

This has been addressed. Current version is:

gist:hasNavigationalParent
	a owl:ObjectProperty ;
	skos:definition "Relates a child category to a parent category in an informal (e.g., faceted) hierarchy."^^xsd:string ;
	skos:example "Refrigerator handles are not refrigerators, but it may be useful to represent their relationship hierarchically for a faceted UI filter."^^xsd:string ;
	skos:prefLabel "has navigational parent"^^xsd:string ;
	.