– DRAFT –
RDF-star Semantics TF

<Souri> Only five of us are on zoom. Were there many regrets for this TF meeting?

<tl> Ora said yesterday that he won't be able to join

<AndyS> will be along in a moment

<TallTed> no regrets on the mailing list

<niklasl> Something to think about: Should there be an OWL class that rdf:reifies belongs to? (Neither owl:DatatypeProperty, nor owl:ObjectProperty?)

<Souri> +1 to "reifier" for the subject of rdf:reifies

<TallTed> rdf:reifies, Domain "reification domain", Range "reification range".

<niklasl> Google result: https://www.yourdictionary.com/reifier > "One who reifies."

<Souri> We do need a name for the "subject" used in a rdf::reifies statement.

<Souri> We may not need to talk frequently about the entity who asserted the rdf:reifies statement. Given that, I like reifier as a way of referring to the subject.

Continue discussion on normal form and semantics

<niklasl> https://www.yourdictionary.com/reification > "The consideration of an abstract thing as if it were concrete, or of an inanimate object as if it were living." So, "concretization"?

<niklasl> <liz> :spouse <richard> . <richard> :spouse <liz> . # Both are reified by the marriage (an event, the more concrete way of conceiving the relationship)?

<niklasl> Or, in OWL, is rdf:reifies an owl:FunctionalProperty?

<niklasl> (Example above implies that it should not be.)

<Souri> <reifier same or different?, tterm same or different?> => four possibilities, if we enumerate. We should think of possible ways of interpreting or managing all of these possibilities (eg., when merging).

<niklasl> No one wants rdf:reifies to be an inverse functional property... (AFAICS)

<niklasl> ex:TripleToken rdfs:subClassOf [ a owl:Restriction ; owl:onProperty rdf:reifies ; owl:cardinality 1 ] . # A triple token is effectively functional in relation to the triple term.

<niklasl> A graph is a set of triples, but a set of triples is not *necessarily* a graph.

<TallTed> enrico noted in zoom chat --

<TallTed> << :b1 | :enrico :born-in :rome >> :on-date 1962 .

<TallTed> << :b1 | :enrico :born-on 1962 >> :location :rome .

<Souri> It should be sufficient to discuss these two (not all four) of the enumerations, right? 1) :e rdf:reifies <<( :s1 :p1 :o1 )>>, <<( :s2 :p2 :o2 )>> . 2) :e1 rdf:reifies <<( :s :p :o )>>, :e2 rdf:reifies <<( :s :p : o )>>.

<enrico> << :b1 | :enrico :born-in :rome >> :on-date 1962 .

<enrico> << :b1 | :enrico :born-on 1962 >> :location :rome .

<Souri> An interesting case=> :e rdf:reifies <<( :john :SSN 123456789 )>>, <<( :john :SSN 234567891 )>> . (and :SSN is supposed to be unique for a person).

<niklasl> Then :e a :Fallacy ?

<niklasl> The domain layer would still allow this, yes.

<Souri> I was thinking of <| ... |> as an alternative. Would that work?

<niklasl> This is would work too <{ :s :p :o }> *grin*

<doerthe> Thank you for generating the minutes Ted, I was late and wanted to catch up :)

<doerthe> we dropped keywords in N3 at some point

<tl> doerthe: why?

<TallTed> EBCDIC

<AndyS> pfps: https://en.wikipedia.org/wiki/UTF-EBCDIC