Re: Suggestions for ACTION-227: Work with eric s on writing section on evolution of duv wrt reuse of namespaces etc.

Hi Antoine,

Berna and I won't meet until Thursday.  I thought I'd make an attempt at a
vocabulary reuse example.  Does this sound appropriate?  The first example
rev:Feedback is where rev:Feedback is not widely known and we wanted a
class to inherit from oa:Annotation.  The second is an argument for
vocabulary reuse because the term already exists and it is widely used.

Eric S

Examples:

Since the beginning, the Data Usage Vocabulary developed by the DWBP
working group has re-used existing vocabularies as much as possible,
following <a href="#ReuseVocabularies">Best Practice 17: Reuse
vocabularies</a>. However, some initial choices of classes and properties
has been revised over the course of vocabulary design. For example although
@@rev:Feedback (http://vocab.org/review/#Feedback)@@ existed, it has been
replaced by a newly defined DUV class @@duv:UserFeedback@@ so that
@@duv:UserFeedback@@ could inherit from @@oa:Anotation@@. The
@@duv:publisher@@ has been replaced by a more commonly used@@dct:publisher@@.
While the constructs initially chosen had appropriate semantics, some
external factors had to be considered. Especially, vocabularies that have
been created and by large groups or consortia and that benefit from their
constant stewardship are more authoritative that vocabularies maintained by
only by one person. Vocabularies created by practitioners from the industry
are likely to be less subject to change than artefacts created by
researchers. A vocabulary hosted on a W3C namespace, like DCAT, PROV or
SKOS, benefits from solid technical guarantees of persistence, as opposed
to a vocabulary hosted on a project's internet domain.

On Tue, Mar 22, 2016 at 12:25 AM, Eric Stephan <ericphb@gmail.com> wrote:

> Hi Antoine,
>
> I plan to meet with Berna either today or Thursday this week.  As part of
> our discussion I am hoping to select representative classes and decisions
> we made to best depict vocabulary reuse.
>
> Cheers,
>
> Eric S
>
> Sent from my iPhone
>
> > On Mar 21, 2016, at 11:28 PM, Antoine Isaac <aisaac@few.vu.nl> wrote:
> >
> > Hi Ghislain,
> >
> > Yes I agree that this should be more specific to the modeling decisions
> of DUV.
> > And the direction you highlight seems good. But before moving on, I'm
> counting on the DUV editors to pick their prefered case of class or
> property selection.
> > This could influence the second part of the paragraph, and also where
> the whole thing stands in the document. I had foreseen it could be an
> example in BP 16, 17 or 18, next to the numbered examples already there
> [1]. It really depends on the case: by default BP17 is the most natural,
> but one could select a class instead of another because it's part of a
> standard (then this would be an example for BP16), or because the class
> initially chosen was too much semantically constrained (then this would be
> an example for BP18).
> >
> > cheers,
> >
> > Antoine
> >
> > [1] http://w3c.github.io/dwbp/bp.html#dataVocabularies
> >
> >> On 3/17/16 3:15 AM, Ghislain Atemezing wrote:
> >> Hi Antoine, all,
> >>
> >> I thought this should be more specific to modeling decisions of DUV.
> Maybe I am wrong here since I’m also missing the context of the action.
> >> I really like the first part of your statement that I copy here
> >>
> >> [
> >>
> >>> Since the beginning, the Data Usage Vocabulary developed by the DWBP
> working group has re-used existing vocabularies as much as possible,
> following <a href="#ReuseVocabularies">Best Practice 17: Reuse
> vocabularies</a>. However, some initial choices of classes and properties
> has been revised over the course of vocabulary design. For example @@Class1@@
> has been replaced by @@Class2@@ and @@prop1@@ has been replaced by
> @@prop2@@
> >> ]
> >>
> >>  as it points out 2 things:
> >>    1- A relation to an existing BP, here BP 17
> >>    2- Some modeling decisions…why class X from domain X is preferred
> instead of class Y from domain Y.
> >>
> >> However, the second part is more generic and could be a BP on criteria
> to reuse vocabularies, part of [1] as referenced in BP 17.
> >> Maybe to be specific to DUV, we can point here why class X  was
> preferred to class Y, based on the general criteria you perfectly mention
> in the second part of your proposal.
> >>
> >> So, we could add the reasons directly of the choices, and to follow on
> your idea, something like:
> >>
> >> [
> >> */For example @@Class1@@ has been replaced by @@Class2@@ because the
> namespace at {domain X} benefits from solid technical guarantees compared /*
> >> */@@prop1@@ has been replaced by @@prop2@@ because the vocab was
> created by practitioners from the industry /*
> >> */]/*
> >>
> >> WDYT?
> >>
> >> Cheers,
> >> Ghislain
> >>
> >> [1] https://www.w3.org/TR/ld-bp/#VOCABULARIES
> >>
> >>> Le 17 mars 2016 à 09:55, Antoine Isaac <aisaac@few.vu.nl <mailto:
> aisaac@few.vu.nl>> a écrit :
> >>>
> >>> Hi Eric, everyone,
> >>>
> >>> Here's a first suggestion for a paragraph on evolution of duv wrt
> reuse of namespaces etc.
> >>> We could add it in one of the vocabulary BPs.
> >>> I'm actually unsure now what was the original context for the action.
> >>> Please tell me if this is alright!
> >>>
> >>> And if you could fill the blanks for the example it would be very
> helpful too as I'm now leaving for a trip...
> >>>
> >>> [
> >>> Since the beginning, the Data Usage Vocabulary developed by the DWBP
> working group has re-used existing vocabularies as much as possible,
> following <a href="#ReuseVocabularies">Best Practice 17: Reuse
> vocabularies</a>. However, some initial choices of classes and properties
> has been revised over the course of vocabulary design. For example @@Class1@@
> has been replaced by @@Class2@@ and @@prop1@@ has been replaced by
> @@prop2@@. While the constructs initially chosen had appropriate
> semantics, some external factors had to be considered. Especially,
> vocabularies that have been created and by large groups or consortia and
> that benefit from their constant stewardship are more authoritative that
> vocabularies maintained by only by one person. Vocabularies created by
> practitioners from the industry are likely to be less subject to change
> than artefacts created by researchers. A vocabulary hosted on a W3C
> namespace, like DCAT, PROV or SKOS, benefits from solid technical
> guarantees of
> >>> persistenc
> >>> e, as opposed to a vocabulary hosted on a project's internet domain.
> >>> ]
> >>>
> >>>
> >>> Cheers,
> >>>
> >>> Antoine
> >>>
> >>>
> >>>> On 2/5/16 3:27 PM, Data on the Web Best Practices Working Group Issue
> Tracker wrote:
> >>>> dwbp-ACTION-227: Work with eric s on writing section on evolution of
> duv wrt reuse of namespaces etc.
> >>>>
> >>>> http://www.w3.org/2013/dwbp/track/actions/227
> >>>>
> >>>> Assigned to: Antoine Isaac
> >
>