W3C home > Mailing lists > Public > public-xg-lld@w3.org > August 2011

RE: References to "application profiles"

From: Young,Jeff (OR) <jyoung@oclc.org>
Date: Thu, 25 Aug 2011 13:37:32 -0400
Message-ID: <52E301F960B30049ADEFBCCF1CCAEF590D928397@OAEXCH4SERVER.oa.oclc.org>
To: "Antoine Isaac" <aisaac@few.vu.nl>, <public-xg-lld@w3.org>
> As far as the "linking issue" appendix is concerned, as discussed
> during the call, I implemented the following modifications:
>
http://www.w3.org/2005/Incubator/lld/wiki/index.php?title=Draft_Vocabul
> aries_Datasets_Section2&diff=5906&oldid=5882

I like this. I'm a little uncomfortable with the overlap between
"semantic alignment" and what the OWL guide calls "ontology mapping":

http://www.w3.org/TR/2004/REC-owl-guide-20040210/#OntologyMapping

I'm sure there is enough of a difference, though, that we shouldn't fret
about it.

> The point was to hint here at some "interoperability gradient".
> Ideally, there would be APs directly re-mixing existing classes and
> properties. 

I'm uncomfortable using the term AP even for situations where a variety
of existing classes and properties are getting reused. An OWL ontology
can be created to document this situation, even if no new terms are
being defined there. Granted, you don't see much of that happening yet,
but presumably that's presumably because people are still trying to wrap
their minds around OWL in general.

> But there are many situations in which re-using a
> vocabulary comes with risks/costs that could motivate coining one's
own
> "duplicate" elements. Consider what schema.org did: they just prefered
> to coin all the elements they need, rather than spend time shopping
> around for existing elements--which may be not well maintained anyway.
> That's maybe not the best practice around, but that will continue to
> happen. In such cases, establishing alignments between element sets is
> a lesser evil.

As a species, I think we are still beginners in terms of modular
thinking. I can't think of a single existing model or set of models that
doesn't make me squirm and shake. "Alignments" or "mappings" give me
hope that we will be able to adapt regardless.

> 
> The "gradient" of best practice here would be: (1) own duplicate
> element sets with no alignment; (2) own element sets with alignment to
> existing vocabularies (3) "true" APs with re-use of existing
> vocabularies.

I agree with these idealized gradients with a couple of caveats. 

1) Existing vocabularies that don't conform to best practice recipes
suck: http://www.w3.org/TR/swbp-vocab-pub/
2) Some existing vocabularies (even popular ones) can be very weird and
arguably would be better off being mapped
3) The ability to transform domain-specific vocabularies into foreign
vocabularies shouldn't be that far off if the mappings exist

> 
> I'm not saying that it was clearly worded here, far from it :-)
> I can also live with this point being mentioned in another section.
But
> I wanted to warn against making this disappear, altogether.

I suspect that all these thoughts are too controversial to used on short
notice, though.

Jeff

> 
> Cheers,
> 
> Antoine
> 
> > Dear all,
> >
> > Re-reading the paragraph in [1]:
> >
> >      A similar concern can be voiced regarding metadata element
sets.
> As
> >      testified in the Linked Open Vocabularies inventory,
> practitioners
> >      generally follow the good practice of re-using existing element
> sets or
> >      building "application profiles" of them. And some projects,
such
> as the
> >      Vocabulary Mapping Framework, aim at supporting that process.
> But the lack
> >      of long-term support for them threatens their enduring meaning
> and common
> >      understanding. Further, some reference frameworks, notably
FRBR,
> have been
> >      implemented in different RDF vocabularies, which are not always
> connected
> >      together. Such situation lowers the semantic interoperability
of
> the
> >      datasets expressed using these RDF vocabularies. The community
> should
> >      encourage the coordinated re-use of element sets for particular
> entity
> >      descriptions, their extension through, e.g., application
> profiles, or their
> >      alignment using, e.g., semantic relations from RDFS and OWL.
> Here, we hope
> >      that better communication between the creators and maintainers
> of these
> >      resources, as encouraged by our own incubator group or the LOD-
> LAM
> >      initiative, will help to consolidate the conceptual connections
> between
> >      them.
> >
> > ...where "a similar concern" refers to "semantic links across value
> vocabularies".
> >
> > Looking closer:
> >
> >> A similar concern can be voiced regarding metadata element sets. As
> >> testified in the Linked Open Vocabularies inventory, practitioners
> >> generally follow the good practice of re-using existing element
sets
> or
> >> building "application profiles" of them.
> >
> > If we mean Dublin-Core-style application profiles (as Singapore
> Framework is
> > cited further on in the paragraph), then we could say something
like:
> >
> >       A similar concern can be voiced regarding metadata element
> sets. As
> >       testified in the Linked Open Vocabularies inventory,
> practitioners
> >       generally follow the good practice of re-using existing
element
> sets or
> >       building "application profiles" that re-use elements from
> multiple sets.
> >
> > Then, I do not really understand the first part of this sentence:
> >
> >> The community should encourage the coordinated re-use of element
> sets for
> >> particular entity descriptions, their extension through, e.g.,
> application
> >> profiles, or their alignment using, e.g., semantic relations from
> RDFS and
> >> OWL.
> >
> > The phrase "encourage the coordinated re-use of element sets for
> > particular entity descriptions" seems to be saying something like:
> >
> >      ...promote the use of common patterns of mixing vocabularies
for
> >      describing particular types of things.
> >
> > However, I do not think this reference to application profiles
really
> belongs
> > in a section on alignment.
> >
> > Rather, I would like to propose the following:
> >
> > -- That the section "The linking issue" (vague, because the whole
LLD
> XG report is
> >     arguably about a "linking issue") be renamed something like:
> >
> >          Semantic alignment
> >
> > -- In this case, the first sentence -- "Many semantic links across
> value
> >     vocabularies are already available..." -- could be preceded with
> a definition
> >     along the lines of:
> >
> >          "Alignments" are links between semantically equivalent,
> similar, or
> >          related terms or entities across different value
> vocabularies, metadata
> >          element sets, or datasets.
> >
> > -- The notion of application profiles is more appropriately
> referenced in the point
> >     about re-using patterns:
> >
> >     In the paragraph:
> >
> >      Design patterns allow implementers to build on the experience
of
> >      predecessors. Traditional cataloging practices are documented
> with a rich
> >      array of patterns and examples, and best practices are starting
> to be
> >      documented for the Linked Data space as a whole (e.g.,
> >      <ref>http://linkeddatabook.com/editions/1.0/#htoc61</ref>). [*]
> What is needed
> >      are design patterns specifically tailored to LLD requirements.
> Such design
> >      patterns would meet the needs of people and developers who
> understand new
> >      techniques through patterns and examples and will increase the
> coherence of
> >      Library Linked Data overall.
> >
> >     I propose inserting a sentence:
> >
> >      Application profiles
(http://dublincore.org/documents/singapore-
> framework/)
> >      provide a method for a community of practice to document and
> share patterns
> >      used for describing specific types of materials.
> >
> > --  ...and application profiles are also relevant to "data design"
> [2]:
> >
> >      Another boost for Linked Data is the growing use of OWL for
> purposes of
> >      data design. Prior to OWL, domain experts could use RDFS to
> create metadata
> >      element sets, but there was no way to map equivalencies across
> >      vocabularies. Among other features, OWL includes an upgrade to
> RDFS to
> >      support ontology mapping. This allows experts to describe their
> domain
> >      using community idioms, while still being interoperable with
> related or
> >      more common idioms. A variety of tools related to OWL can be
> found on the
> >      W3C's RDF wiki and OWL wiki. Unified Modeling Language (UML)
> tools are also
> >      value to help designers represent and manipulate domain models
> visually.
> >      The Ontology Definition Metamodel (ODM) specification should
> help bridge
> >      some of the gaps between UML and OWL. [*]
> >
> >     I propose to add:
> >
> >      Application profiles
(http://dublincore.org/documents/singapore-
> framework/)
> >      provide a way to specify how a community of practice defines a
> >      domain model and re-uses specific vocabularies in order to
> create metadata
> >      conforming to a particular pattern.
> >
> > Tom
> >
> > [1]
>
http://www.w3.org/2005/Incubator/lld/wiki/DraftReportWithTransclusion#T
> he_linking_issue
> > [2]
>
http://www.w3.org/2005/Incubator/lld/wiki/DraftReportWithTransclusion#T
> ools_for_data_designers
> > [3]
>
http://www.w3.org/2005/Incubator/lld/wiki/DraftReportWithTransclusion#D
> evelop_and_disseminate_best-practices_design_patterns_tailored_to_LLD
> >
> >
> 
> 
Received on Thursday, 25 August 2011 17:38:41 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 25 August 2011 17:38:42 GMT