Definition of gist:Message is out of sync with the formal definition

[rjyounes]

Human-readable definition (before update to SKOS annotations):

	rdfs:comment
		"A specific instance of content sent from an Organization, Person, or Application to at least one other Organization, Person, or Application."^^xsd:string ,
		"EXAMPLES: An email message, a phone call, a voice message, or a Web Service message."^^xsd:string
		;

Formal definition:

	owl:equivalentClass [
		a owl:Class ;
		owl:intersectionOf (
			gist:ContentExpression
			[
				a owl:Restriction ;
				owl:onProperty gist:fromAgent ;
				owl:minCardinality "1"^^xsd:nonNegativeInteger ;
			]
			[
				a owl:Restriction ;
				owl:onProperty gist:toAgent ;
				owl:minCardinality "1"^^xsd:nonNegativeInteger ;
			]
		) ;
	] ;
	.

NOTE: This formal definition has been modified in PR #395. See notes there on rationale. Previously it was

	owl:equivalentClass [
		a owl:Class ;
		owl:intersectionOf (
			gist:ContentExpression
			[
				a owl:Restriction ;
				owl:onProperty gist:fromAgent ;
				owl:someValuesFrom [
					a owl:Class ;
					owl:unionOf (
						gist:Address
						gist:Organization
						gist:Person
					) ;
				] ;
			]
			[
				a owl:Restriction ;
				owl:onProperty gist:toAgent ;
				owl:someValuesFrom [
					a owl:Class ;
					owl:unionOf (
						gist:Address
						gist:Organization
						gist:Person
					) ;
				] ;
			]
		) ;
	] ;

	.

We removed Address from the ranges of toAgent and fromAgent in issue #391. In coming across the definition of Message, rather than remove Address there as well, I replaced the someValuesFrom restriction with a minCardinality restriction, since the unionOf simply restates the property ranges.

This re-raises the issue of defining a SoftwareAgent or SoftwareApplication class. This is not the only time it is referenced in conjunction with Person and Organization. See, for example, the definition of Event:

Something happening over some period of time, often characterized as some kind of activity being carried out by some person, organization, or software application.

Should we have a class SoftwareAgent or SoftwareApplication to be added to the ranges of fromAgent and toAgent?

This question was raised previously in the closed issue #132, where the addition of an Agent class was rejected. But we did not deal in depth with the issue of software agents. We should either revisit this or adjust the human-readable definitions of Message, Event, and any others, accordingly.

[uscholdm]

Should we have a class SoftwareAgent or SoftwareApplication to be added to the ranges of fromAgent and toAgent?

I'm inclined to say yes; if we do, we should use whatever is in the software ontology.

BTW, as of 23-Oct-23, the restrictions on Message again explicitly refer to Organization and Person. See below.

gist:Message
	a owl:Class ;
	rdfs:isDefinedBy <https://w3id.org/semanticarts/ontology/gistCore> ;
	owl:equivalentClass [
		a owl:Class ;
		owl:intersectionOf (
			gist:ContentExpression
			[
				a owl:Restriction ;
				owl:onProperty gist:comesFromAgent ;
				owl:someValuesFrom [
					a owl:Class ;
					owl:unionOf (
						gist:Organization
						gist:Person
					) ;
				] ;
			]
			[
				a owl:Restriction ;
				owl:onProperty gist:goesToAgent ;
				owl:someValuesFrom [
					a owl:Class ;
					owl:unionOf (
						gist:Organization
						gist:Person
					) ;
				] ;
			]
		) ;
	] ;
	.

[sa-bpelakh]

DECISION: Remove type references from definition, e.g

	rdfs:comment
		"A specific instance of content sent from a sender to at least one other recipient."^^xsd:string ,
		"EXAMPLES: An email message, a phone call, a voice message, or a Web Service message."^^xsd:string
		;