should gist:isCharacterizedAs be a subprop of gist:isCategorizedBy?

[justin2004]

I think gist:isCharacterizedAs is a specialization of gist:isCategorizedBy where the domain and range are fixed.
Whereas gist:isCategorizedBy doesn't fix the domain and range.

[uscholdm]

I think gist:isCharacterizedAs is a specialization of gist:isCategorizedBy

If anything, it is the reverse. One way to characterize something is to say what category it is in. There are possibly many other ways. However, we might not need gist:isCharacterizedAs - I've never found a use case for it.

Ironically, gist:isCharacterizedAs is uncharacteristic of most gist properties, by having a very narrow domain and range. That goes against our guidelines. If we keep gist:isCharacterizedAs I hitnk it shoujld have no domain or range and should be a supeproperty of gist:isCategorizedBy.

[justin2004]

I think it should have no domain or range and should be a supeproperty of gist:isCategorizedBy.

I'd be happy with that as well, @uscholdm .
I think they likely belong in the same prop hierarchy (not top level siblings).

I was just using the presence of domain and range to be a cue that this is a specialization. Also isCharacterizedAs has "categorize" in its skos:defintion.

Also the skos:definition on gist:isCharacterizedAs seems to be backwards:
"A way to categorize a behavior."
While the domain is gist:Event and the range is gist:Behavior.
So I'd say that the Event is categorized by the Behavior -- the Behavior isn't being categorized.

[rjyounes]

I agree, the current definition seems a complete anomaly. I'd be in favor of dropping it, as you initially suggest, @uscholdm. What is the use case when isCategorizedBy is not appropriate?

[uscholdm]

Also the skos:definition on gist:isCharacterizedAs seems to be backwards:
"A way to categorize a behavior."

That's true. This is an ancient definition, needs to be updated if we keep the property.

[uscholdm]

PROPOSAL: delete the property isCharacterizedAs because it is historical and was never used, to my knowledge. The narrow domain and range are also historical, we don't do that these days. The main purpose it serves is to enable Event to have some axioms. That is not a super-compelling reason.

gist:Event
	a owl:Class ;
	owl:equivalentClass [
		a owl:Restriction ;
		owl:onProperty gist:isCharacterizedAs ;
		owl:someValuesFrom gist:Behavior ;
	] ;


gist:isCharacterizedAs
	a owl:ObjectProperty ;
	rdfs:domain gist:Event ;
	rdfs:range gist:Behavior ;
	owl:deprecated "true"^^xsd:boolean ;
	skos:definition "A way to categorize a behavior."^^xsd:string ;
	skos:prefLabel "is characterized as"^^xsd:string ;
	.

[rjyounes]

We have deprecated the property in PR #951, so the issue is irrelevant.