Re: Basic schema.org / RDFa Lite library catalog for your amusement / derision

Karen: 

Many thanks for your note - in fact, I somehow completely missed
http://schema.org/MusicAlbum when I originally set up the
schema.org mapping and jumped (invalidly, it appears) directly to
MusicRecording, so it sounds like I have some fixing to do. (I'd love to
claim that MusicAlbum didn't exist back when I first implemented this,
but I'm sure that a version control system would show that I'm just a
bonehead and didn't dig deep enough back in January). 

Now that you mention it, I guess the 700 ind2=2 is the key for
differentiating the MusicRecording objects that should be inAlbum, vs.
plain old added entries for authors and the like. Also, right now I
believe I have Evergreen just pulling in the 245 blindly, so adding some
smarts around making the uniform title take precedence over the 245
would make a lot of sense. 

If I get a few minutes tonight I'll take another stab at improving
this. Collaboration and iteration for the win (hopefully). 

Thanks again, Karen! 

Dan

>>> Karen Coyle <kcoyle@kcoyle.net> 12/5/2012 5:08 PM >>>
Dan, I haven't had a chance to look at these yet but will definitely
do
so. Just to say that I had tasked myself with creating some music
examples, and ended up stumped on my very first attempt:

LC Control No.: 99593299
LCCN Permalink: http://lccn.loc.gov/99593299
000 01227cjm a22003132a 450
001 12077846
005 20060214161519.0
007 sdubsmennmplue
008 870915s1964 xx con d
035 __ |a (OCoLC)ocm34455432
010 __ |a 99593299
028 02 |a 6880 014 |b Philips
040 __ |a CUT |c CUT |d OCL |d DLC
042 __ |a lcderive
050 00 |a Philips 6880 014
100 1_ |a Grieg, Edvard, |d 1843-1907.
240 10 |a Concerto, |m piano, orchestra, |n op. 16, |r A minor
245 00 |a Piano concerto in A minor, op. 16 |h [sound recording] / |c
Edvard Grieg. Piano concerto in A minor, op. 54 / Robert Schumann.
260 __ |a [S.l.] : |b Philips, |c [1964?]
300 __ |a 1 sound disc : |b analog, 33 1/3 rpm. stereo. ; |c 12 in.
511 0_ |a Claudio Arrau, piano ; Concertgebouw-Orchestra, Amsterdam ;
Christoph von Dohnányi, conductor.
500 __ |a Robert Oursler donation.
650 _0 |a Concertos (Piano)
700 12 |a Schumann, Robert, |d 1810-1856. |t Concertos, |m piano,
orchestra, |n op. 54, |r A minor.
700 1_ |a Dohnányi, Christoph von. |4 cnd
700 1_ |a Arrau, Claudio, |d 1903-1991. |4 prf
710 2_ |a Concertgebouworkest. |4 prf

The first 700 could get an "/inAlbum" property, but the other 700's do
not. And the 245 isn't a name for the album it's for one of the pieces
in the album. I think I just over-think this kind of case. Anyway,
would
love to see what Evergreen does with these.

kc

On 12/5/12 7:40 AM, Dan Scott wrote:
> Hey folks:
>
>
> One of my action items, if I recall correctly, was to provide a
publicly
> accessible catalog publishing schema.org microdata and RDFa Lite
based
> on in-the-wild library data.
>
>
> Therefore, see the following:
>
>
> * schema.org: _/http://laurentian-test.concat.ca/_ (this is pretty
much
> what you get from Evergreen for schema.org out-of-the-box - such as
it is)
>
> * RDFa Lite: _/http://rdfa_lite_test.concat.ca/_ (this is a naive,
fast
> as possible variation on the Evergreen schema.org microdata that
uses
> the schema.org vocabulary and nothing more)
>
>
> The microdata is not exposed at any level other than the record
details
> page, so you can search for your favourite kind of record and drill
down
> from there to compare. For example:
>
>
> Book: _/http://laurentian-test.concat.ca/eg/opac/record/858351/_ vs.
> _/http://rdfa_lite_test.concat.ca/eg/opac/record/858351/_
>
> MusicRecording:
> _/http://laurentian-test.concat.ca/eg/opac/record/854855/_ vs.
> _/http://rdfa_lite_test.concat.ca/eg/opac/record/854855/_
>
>
> Oh, that's actually it for mappings from MARC to schema.org
> CreativeWorks at this point. I'd be happy to model more for notated
> music, etc, based on the group's suggestions, but I'm starting basic
and
> working up from there.
>
>
> I've done some tests with Google's rich snippets testing tool (
> _/http://www.google.com/webmasters/tools/richsnippets/_ ), the w3c
Nu
> validator ( _/http://validator.w3.org/nu/_ ), and the RDFa extractor
(
> _/http://getschema.org/index.php/Main_Page/_ ). The Nu validator, as
you
> would expect, is grumpy about violations of even the HTML5 markup
but
> doesn't seem to complain about the RDFa Lite implementation; the
other
> tools seem quite happy with both the schema.org and RDFa Lite
versions.
>
>
> Some notes about the underlying data:
>
> * Most of the records were copy catalogued from a variety of
sources,
> but there is some original cataloguing in the mix.
>
> * The data is from a mid-sized bilingual academic institution and
> reflects a fairly broad range of subjects and materials. Yes, we
have
> catalogued realia (puppets for our education school).
>
>
> I've stored the RDFa Lite implementation for Evergreen in git at
>
_/http://git.evergreen-ils.org/?p=working/Evergreen.git;a=shortlog;h=refs/heads/user/dbs/rdfa_lite/_
-
> currently the top commit is all that I needed to get the (very
basic)
> version rolling, as it's a light customization layer on top of the
base
> Evergreen skin. We use Perl's Template::Toolkit as the basis of our
> catalogue web UI, in case anyone wants to take a shot at taking this
> further. The Open-ILS/src/templates/opac directory contains the base
> layer for the catalogue, and the Open-ILS/src/templates_rdfa
directory
> overlays its templates when you visit rdfa_lite_test.concat.ca. If I
had
> more time, I could & would have abstracted things out a bit so that
> switching from schema.org to RDFa Lite was just a matter of throwing
a
> config switch rather than touching a handful of templates. Oh well!
>
>
> Apologies overall for the hastiness of the implementation and just
> throwing this semi-structured email onto the list; this week I'm in
> lockdown for 12 hours a day at a documentation sprint (thanks,
Google
> and FLOSS Manuals) which isn't leaving much time or brainpower for
> anything else.
>
>
> Thanks,
>
> Dan
>

--
Karen Coyle
kcoyle@kcoyle.net http://kcoyle.net
ph: 1-510-540-7596
m: 1-510-435-8234
skype: kcoylenet