Re: Use Cases for Choice, Composite, and List

Actually the Digital Publishing Interest Group's use cases, published as a
note here:
    http://www.w3.org/TR/dpub-annotation-uc/
Which maybe adds more weight :)

We discussed dropping composite at last year's TPAC and the decision was to
keep it.  We have not discussed it since then, to my recollection.

Rob


On Fri, Oct 2, 2015 at 8:42 AM, Jacob Jett <jjett2@illinois.edu> wrote:

> Hi all,
>
> Doug asked about the use cases for the multiplicity types in the current
> draft of the model. A well developed set of these use cases (sans examples
> in json or ttl) can be found in the Community Group's dpub use cases
> document: http://w3c.github.io/dpub-annotation/
>
> A more detailed explanation for where these three things (Choice,
> Composite, and List) originated from is as follows.
>
> Section 5 (Mulitplicity) of the data model arose out of a consesus in the
> Community Group to accommodate several use cases that were developed during
> the early OAC work. A discussion of them can be found in the Community
> Group Archives.[1]
>
> The use case for oa:List emerged from work on annotating videos that took
> place at MITH (U. Maryland). It quickly emerged that multiple selectors
> would be necessary in order for specific time-space regions of videos to be
> annotated. This issue appears (somewhat under-developed) in the OACG
> wiki[2] and MITH's final report (where an earlier attempt at accommodating
> it (oa:CompositeSelector) is illustrated.[3] During one of the early CG
> meetings it was pointed out that you wouldn't apply the selectors at the
> same time (i.e., "composite" has semantic senses other than that of list or
> ordered list) and so oa:List emerged.
>
> Example 63 in the current FPWD showcases the expected pattern for
> oa:List.[4]
>
> {
>   "@id": "http://example.org/anno1",
>   "@type": "oa:Annotation",
>   "body": {"@id": "http://example.org/body1"},
>   "target": {
>     "@id": "http://example.org/sptarget1",
>     "@type": "oa:SpecificResource",
>     "source": "http://example.org/target1",
>     "selector": {
>       "@type": "oa:List",
>       "members": [
>         "http://example.org/selector1",
>         "http://example.org/selector2"
>       ]
>     }
>   }
> }
>
> Another place where this might be used is in the editing use case. Imagine
> a single user who creates a batch of edits through a single annotation
> (IIRC this is similar to Bob Morris's Filtered Push and Lutz's use cases).
>
> {
>   "@id": "http://example.org/anno1",
>   "@type": "oa:Annotation",
>   "motivation": "editing",
>   "body": {
>     "@id": "http://example.org/list_body1",
>     "@type": "oa:List",
>       "members": [
>         "http://example.org/edit1",
>         "http://example.org/edit2",
>         "http://example.org/edit3",
>         "http://example.org/edit4",
>         "http://example.org/edit5",
>   },
>   "target": {"@id": "http://example.org/target1"}
> }
>
> There several ways of making this complex annotation even moreso, all
> having to do with adding additional bodies or members with different roles.
> I'll defer those to a later date (hopefully much later).
>
> List's sister oa:Composite was (as noted above) was the first attempt to
> provide something unordered. Unfortunately there's really no such thing as
> an unordered list, they just come with an arbitrary order or an enumerated
> order. Some early deliberations of the OACG suggested that it be eliminated
> altogether[5]
>
> The notion of composites was re-developed to satisfty a use case that came
> out of OAC work here at Illinois. Scholars' need to be able to forge a
> single target comprising two juxtaposed images [6] (see especially section
> 4.3 of [6]) was the driving use case for the development of oa:Composite.
> And the use case fully developed and documented in the OACG's electronic
> publishing oriented use case document[7] at [8] and [9] respectively
> (although there are no illustrative examples in this document).
>
> The current FPWD lays out the expected pattern with example 61
> (below).[10]
>
> {
>   "@id": "http://example.org/anno1",
>   "@type": "oa:Annotation",
>   "body": {"@id": "http://example.org/body1"},
>   "target": {
>     "@type": "oa:Composite",
>     "item": [
>       "http://example.org/target1",
>       "http://example.org/target2"
>     ]
>   }
> }
>
> The notion of oa:Choice emerged similarly to that of oa:List. In the
> beginning it was intended to operate with Selectors. It quickly became
> apparent during the face-to-face meetings that it also had a valuable
> purpose in the context of bodies and targets, the generalities of which are
> recorded in OACG wiki[11] and the OACG use case document[7], especially at
> section 2.3. [12]
>
> The current FPWD lays out the expected pattern in example 59 (below).[13]
>
> {
>   "@id": "http://example.org/anno1",
>   "@type": "oa:Annotation",
>   "body": {
>     "@type": "oa:Choice",
>     "members": [
>       "http://example.org/body1",
>       "http://example.org/body2"
>     ]
>   },
>   "target": "http://example.org/target1"
> }
>
>
>
> 1:
> https://lists.w3.org/Archives/Public/public-openannotation/2012Oct/0004.html
> 2: https://www.w3.org/community/openannotation/wiki/Issue_lists
> 3:
> http://openannotation.org/documents/ProjectReports/MarylandFinalReport.pdf
> 4: http://www.w3.org/TR/annotation-model/#list
> 5: https://www.w3.org/community/openannotation/wiki/Issue_composite
> 6:
> http://openannotation.org/documents/ProjectReports/EmblematicaFinalReport.pdf
> 7: http://w3c.github.io/dpub-annotation/
> 8:
> http://w3c.github.io/dpub-annotation/#annotation-comparing-segments-within-a-publication
> 9:
> http://w3c.github.io/dpub-annotation/#annotation-comparing-segments-between-publications
> 10: http://www.w3.org/TR/annotation-model/#composite
> 11: https://www.w3.org/community/openannotation/wiki/Issue_choicelists
> 12: http://w3c.github.io/dpub-annotation/#advanced-annotation-use-cases
> 13: http://www.w3.org/TR/annotation-model/#choice
>
>
>
>
> Regards,
>
> Jacob
>
> _____________________________________________________
> Jacob Jett
> Research Assistant
> Center for Informatics Research in Science and Scholarship
> The Graduate School of Library and Information Science
> University of Illinois at Urbana-Champaign
> 501 E. Daniel Street, MC-493, Champaign, IL 61820-6211 USA
> (217) 244-2164
> jjett2@illinois.edu
>
>


-- 
Rob Sanderson
Information Standards Advocate
Digital Library Systems and Services
Stanford, CA 94305