Re: Dividing up "Examples Library" for schemabibex?

Hi Dan,

I agree with many of the points you make.

Pragmatically, as I wanted the navigation links between the examples in
the library to actually work, I went for the solution of an imaginary
library which hosted everything.  In reality I would expect the norm to
probably be as you suggest 'the Work / Expression level records live in an
"Authoritative Bibliographic Resource" web site (Freebase, Wikipedia,
WorldCat, OpenLibrary, what have you)'

The effect of removing many of the "hasInstanceOf" from the higher level
records is also probable where the authoritative site does not have
knowledge of individual library resources.  Many consortial or aggregation
sites may well have this information however so it would not be too
unusual to find such properties populated.

~Richard.

On 08/04/2013 19:29, "Dan Scott" <dscott@laurentian.ca> wrote:

>Hi all:
>
>I just added 
>http://www.w3.org/community/schemabibex/wiki/Talk:Examples/mylib to the
>overall Examples Library section. In essence, my concern is that library
>catalogues are unlikely to surface user-visible (Web crawlable) records
>at the Work / Expression levels, and that keeping all of the examples
>with the "Examples Library" doesn't match what most libraries are likely
>to experience in their attempts to add schema.org structured data to
>their catalogues. (Substitute "catalogue developers" accordingly).
>
>Therefore I've suggested in the Talk section that we make the Work /
>Expression level records live in an "Authoritative Bibliographic
>Resource" web site (Freebase, Wikipedia, WorldCat, OpenLibrary, what have
>you), and have the "Examples Library" play the role of a public,
>academic, or special library... in which case, they would simply link
>their Manifestation / Instance records to the authoritative resource.
>
>Come to think of it, that may also remove most of the "hasInstanceOf"
>properties from the higher level records that point to the lower-level
>records; we could comment where that might occur (for example, in a
>library catalogue that does surface Work / Expression records, or in
>WorldCat), but I think that would make for a more realistic, broader
>scenario on both the higher-level and lower-level records. I know that
>the current arrangement was done for perceived simplicity, but I suspect
>that it might actually detract from the usability of the examples for
>implementers.
>
>Dan
>
>
>

Received on Monday, 8 April 2013 21:57:49 UTC