Re: Updated Schema Architypes Straw Man Proposal

Hi Elizabeth,

Thank you for this very useful contribution.

Let me first try to answer the points raised from your meeting.

·         We’re unclear what the difference between the URL and the @id
> property is.  If we have a URI for the collection, which schema property
> should we be using?
>
This slightly confusing area is a function of Schema.org being widely
acceptable in three serialisation syntaxes - Microdata, RDFa, JSON-LD.  As
Microdata and RDFa are embedded within the html attributes of a page, the
identifier for the described entity is by default assumed to be the url of
the page the data is embedded within, unless that is overridden by
supplying a specific url property.  This often when more than one entity,
or Thing, is described on a single page.

Using  JSON-LD syntax in a self-contained data element that is inserted
within a page [to share with search engines] or a standalone file, the
convention is to provide an identifier for an entity in the @id property.
So in this case the inclusion of a url property would be superfluous.

Increasingly JSON-LD is becoming the norm for sharing Schema.org,
especially since Google and others use it for their examples, so using @id.


> I’ll also mention that I’m still a bit confused about the URIs we’re
> using.  In the examples Richard posted, the URIs are URIs for the finding
> aid description of the collection, not the collection/object itself.
> Should we make that clear somehow using a /#collection convention or
> something else?
>
One has to take care in assigning semantic meaning to URIs.  They are just
and identifier for a Thing, it is the properties of that Thing that should
define what type of thing it is.  The construction of a URI and its path
elements are more to do with the supporting technology than the things
given those URI identifiers.  It is precisely for this reason opaque URIs
are often recommended (e.g.. http://myarchive.org/aqz3459wpt) to prevent
wrong semantic assumptions being made by the consumers of data.

So to that end, meaning should be only inferred from the Schema.org Type(s)
used to describe an entity (Archive, ArchiveCollection, ArchiveItem,
CreativeWork, etc. or combination thereof).



> ·         We’d like a more specific definition of what an ArchiveItem
> is.  Archival description being as hierarchical as it is, we can’t figure
> out when we’d use ArchiveItem except perhaps for individual digitized
> objects.  Usually the lowest level in an archival description is a folder,
> which would be a collection of ArchiveItems.
>
Note from the description of ArchiveItem
<http://archive.sdo-archive.appspot.com/ArchiveItem> it has *isPartOf* and
*hasPart* as available properties.  So for example a box of stuff could be
described as an *ArchiveItem* that potentially *hasPart*(s) which are also
individual *ArchiveItem*s. To push the analogy further, some of those items
could be folders that themselves contain items with may or may not be
described.

In the vocabulary world of Schema.org assigning a Type is more about making
available sets of descriptive properties, than in an ontological one, where
assigning a class can assert behaviour.   It is more than possible to
define something to be both an *ArchiveType* and an *ArchiveCollection*.

Having said that, the descriptions of proposed terms are as draft as the
model its self - suggestions welcome.



> ·         We agree with others who have said that we need an extent
> property that conveys the size of the ArchiveCollection.  We’d want to use
> this at the collection/subcollection/item levels.
>

Ok. What we need therefore is to get some consensus around what we would
call that property and some guidance on how it would be used.  (See my
previous comments on this)

>
>
> We’ve also done some mappings between schema and DACS/ISAAR-CPF/ISAD(G)
> and two archival management systems (ArchivesSpace and AtoM).  These are
> still a work-in-progress, but are open if anyone is interested:
> https://docs.google.com/spreadsheets/d/1jsPRML3BCkF4EdWTR4r2ooAw2iTnd
> ErLZlse0cs3Osk/edit#gid=0 .
>


 I will take a separate look at this but it looks like a good start - one
instant comment: The default property for names should be *name*, using
*givenName* etc. to provide extra detail if available.


> Our impetus for doing so is that we hope to both use schema for front end
> interfaces as well as outputting archival description in LOD.  We realize
> that we will have to use other ontologies to fully express all of our
> metadata in LD, but we’re trying to see how far we can get with schema due
> to wide adoption of the namespace.
>

Exactly the right approach to take with Schema.org!  It is a general
purpose vocabulary for sharing data with a wide cross domain audience, it
is unlikely it will totally fit a single detailed domain but is often
surprising how far you can get with it.

~Richard.

>
>
> Elizabeth
>
>
>
> ___________________________
>
> Elizabeth Russey Roke
>
> Digital Archivist
>
> Stuart A. Rose Manuscript, Archives, & Rare Book Library
>
> 404.727.2345 <(404)%20727-2345> | erussey@emory.edu
>
>
>
>
>
>
>
> "The Stuart A. Rose Manuscript, Archives, & Rare Book Library collects and
> connects stories of human experience, promotes access and learning, and
> offers opportunities for dialogue for all wise hearts who seek knowledge.”
>
>
>
> Read the Rose Library blog: https://scholarblogs.emory.edu/marbl/
>
>
>
> Like the Rose Library on Facebook:  https://www.facebook.com/emorymarbl
>
>
>
> Follow the Rose Library on Twitter: https://twitter.com/EmoryRoseMARBL
>
>
>
> *From: *Richard Wallis <richard.wallis@dataliberate.com>
> *Date: *Thursday, May 18, 2017 at 11:43 AM
> *To: *Owen Stephens <owen@ostephens.com>
> *Cc: *Jane Stevenson <Jane.Stevenson@jisc.ac.uk>, public-architypes <
> public-architypes@w3.org>
> *Subject: *Re: Updated Schema Architypes Straw Man Proposal
> *Resent-From: *<public-architypes@w3.org>
> *Resent-Date: *Thursday, May 18, 2017 at 11:44 AM
>
>
>
> To answer both Jane and Owen….
>
>
>
> The semantics (no pun intended) I was trying to convey in my possibly
> contrived example was this:
>
>    - The *AudioObject* *Sound recording of … My Grandfather’s Forehead ..*
>
>
>    - Has a URI of https://archiveshub.jisc.ac.uk/data/gb71-thm/407/thm/
>       407/8/3
>       - Is an *ArchiveItem*
>       - It is *isPartOf* an *ArchiveCollection* with the URI *https://archiveshub.jisc.ac.uk/data/gb71-thm/407/thm/407/8
>       <https://archiveshub.jisc.ac.uk/data/gb71-thm/407/thm/407/8>*
>
>
>    - There is an *ArchiveCollection* with the URI
>    https://archiveshub.jisc.ac.uk/data/gb71-thm/407/thm/407/8
>    <https://archiveshub.jisc.ac.uk/data/gb71-thm/407/thm/407/8>
>
>
>    - Which contains Sound Recordings about Ronnie Barker
>       - It *hasPart* *https://archiveshub.jisc.ac.uk/data/gb71-thm/407/thm/407/8/3
>       <https://archiveshub.jisc.ac.uk/data/gb71-thm/407/thm/407/8/3>*
>       - It is also *partOf* another ArchiveCollection with the*
>       URI https://archiveshub.jisc.ac.uk/data/gb71-thm/407
>       <https://archiveshub.jisc.ac.uk/data/gb71-thm/407>*
>       - It could be considered as a sub-collection of that other
>       collection
>
>
>    - There is an *ArchiveCollection* with the URI
>    https://archiveshub.jisc.ac.uk/data/gb71-thm/407/thm/407
>    <https://archiveshub.jisc.ac.uk/data/gb71-thm/407/thm/407>
>
>
>    - It contains other collections
>       - It contains an *ArchiveCollection* with the URI
>       https://archiveshub.jisc.ac.uk/data/gb71-thm/407/thm/407/8
>       <https://archiveshub.jisc.ac.uk/data/gb71-thm/407/thm/407/8>
>       - It potentially does not directly contain individual *ArchiveItem*s
>
> In this case the example *ArchiveItem* can be considered to be *partOf*
> the sub-collection <*https://archiveshub.jisc.ac.uk/data/gb71-thm/407/thm/407/8
> <https://archiveshub.jisc.ac.uk/data/gb71-thm/407/thm/407/8>*> and,
> because that is also partOf the main collection, it is also partOf that as
> well.   For identifying and sharing the relationship using structured data
> however, it would not be necessary, and possibly even confusing for
> consumers, to explicitly assert that.
>
>
>
> As I say, my example may be contrived to demonstrate the possibility of
> collections within collections.
>
>
>
> To answer Jane with regard to *archiveHeld* …
>
>
>
> The example indicates that one of the *archiveHeld* (held, kept or
> maintained by) by the *Archive* organization (“V&A Theatre and
> Performance Collections”) is https://archiveshub.jisc.
> ac.uk/data/gb71-thm/407/thm/407 (The Ronnie Barker Collection).
>
>
>
> It also indicates, using the *location* property of *ArchiveCollection*,
> that the Ronnie Barker Collection is located at the “V&A Theatre and
> Performance Collections” Archive.  Although looking like the reverse of a
> similar relationship these two statements are defining different things.
> Imagine for example that the “V&A Theatre and Performance Collections”
> Archive was responsible for the Ronnie Barker Collection, but it was housed
> at the British Museum. In such a case the *location* property becomes
> important.
>
>
>
> Finally, Jane:
>
>  If we did use something like collectionSize then that would imply top
> level (collection level)
>
>
>
> Why would it imply anything other than the size of the *ArchiveCollection*
> being described, be it a sub-collection, a single collection, a
> collection containing sub-collections?
>
>
>
>
>
> ~Richard.
>
>
>
>
>
> On 18 May 2017 at 10:48, Owen Stephens <owen@ostephens.com> wrote:
>
>
>
> On 18 May 2017, at 10:34, Jane Stevenson <Jane.Stevenson@jisc.ac.uk>
> wrote:
>
>
>
> 4. However I concur that in such a case in practice it probably would not
> be practical to list all 500 in the JSON-LD insert on the collection page.
>   In such a case however use of isPartOf In the description of the
> ArchiveItem would be sufficient to assert the relationship to a search
> engine:
> “isPartOf”: “https://archiveshub.jisc.ac.uk/data/gb71-thm/407/thm/407/8”
> (JSON-LD syntax)
>
>
> Yes, that’s exactly my thinking.
>
> But it would be “isPartOf”: “https://archiveshub.jisc.ac.
> uk/data/gb71-thm/407/thm/407” (is part of the Ronnie Barker Collection
>
>
>
> Trying to get this straight in my mind - would you describe the specific
> Item (https://archiveshub.jisc.ac.uk/data/gb71-thm/407/thm/407/8/3) as
> being part of:
>
>
>
> a) the ‘sub-collection’ (https://archiveshub.jisc.ac.
> uk/data/gb71-thm/407/thm/407/8)
>
> b) the archive collection (https://archiveshub.jisc.ac.
> uk/data/gb71-thm/407)
>
> c) both
>
>
>
> ?
>
>
>
> Owen
>
>
>
>
>
> ------------------------------
>
> This e-mail message (including any attachments) is for the sole use of
> the intended recipient(s) and may contain confidential and privileged
> information. If the reader of this message is not the intended
> recipient, you are hereby notified that any dissemination, distribution
> or copying of this message (including any attachments) is strictly
> prohibited.
>
> If you have received this message in error, please contact
> the sender by reply e-mail message and destroy all copies of the
> original message (including attachments).
>