- From: Young,Jeff (OR) <jyoung@oclc.org>
- Date: Mon, 9 Aug 2010 12:39:14 -0400
- To: "Jodi Schneider" <jodi.schneider@deri.org>
- Cc: "Karen Coyle" <kcoyle@kcoyle.net>, <public-xg-lld@w3.org>
- Message-ID: <52E301F960B30049ADEFBCCF1CCAEF59094295F5@OAEXCH4SERVER.oa.oclc.org>
Jodi,
You asked "Do you imagine an agent using that subject?" The simple
answer is yes: I imagine human, machine, and semantic agents "using"
both the concept of World War II and the LCSH subject heading "World
War, 1939-1945". The concept of World War II is useful because people
continue to write "books" on it. The LC Subject Heading "World War,
1939-1945" is useful because it is a network/schematic label/name for
that concept. If Web architectures keep the identify of "the thing"
separate from "the label/name of the thing", then most network users
(e.g. Google searchers) can continue to ignore this subtle difference.
As I understand it, this is where FRSAD/SKOS compete as a solution:
SKOS makes a distinction between owl:Thing and skosxl:Label
<http://www.w3.org/TR/skos-reference/skos-xl.html#Label> as classes and
defines the relationship properties skosxl:prefLabel
<http://www.w3.org/TR/skos-reference/skos-xl.html#prefLabel> or
skosxl:altLabel
<http://www.w3.org/TR/skos-reference/skos-xl.html#altLabel> to connect
them. For example, all the concepts in LCSH are identified as
skos:Concepts <http://www.w3.org/TR/skos-reference/#concepts> , but
beware that the subject heading is NOT the concept. They are (or at
least should be) treated as two different things. If LCSH upgraded their
skos:ConceptSchemes
<http://www.w3.org/TR/skos-reference/skos.html#ConceptScheme> to use
skosxl:prefLabel and skosxl:altLabel instead of skos:prefLabel
<http://www.w3.org/TR/skos-reference/skos.html#prefLabel> and
skos:altLabel <http://www.w3.org/TR/skos-reference/skos.html#altLabel> ,
then this would be clearer. Even if they don't, though, the concept and
the subject headings are still two different things. The difference is
that that in LCSH the skos:Concept ("the thing") is identified with an
HTTP URI but the subject heading ("the label/name of the thing") is not.
If that sounds weird, think closely about SKOS XL:
http://www.w3.org/TR/skos-reference/skos-xl.html.
In FRSAD <http://www.ifla.org/node/1297> , owl:Thing is analogous to
frsad:Thema and skosxl:Label is analogous to frsad:Nomen. For IFLA, the
basic issue seems to have originated as they considered the appropriate
range for the FRBR "has as subject" relationship. FRBR clearly sets the
"Work" as the domain for this relationship, but they never gave a name
to the range class. FRSAD choose the class name "Thema" because this
Latin term carried as little baggage as possible and (theoretically)
includes anything imaginable. They then created a "Nomen" class to
decouple the controlled vocabulary terms and created
frsad:hasAppellation and frsad:isAppellationOf properties to connect
Themas and Nomens. IMO, this is the same thing SKOS XL is trying to do.
The only mentionable difference between the SKOS and FRSAD models is
that in SKOS the "scheme" attaches to "the thing" whereas in FRSAD the
"scheme" attaches to "the name of the thing". The choice seems arbitrary
to me and thus doesn't justify us inventing a library variant of
SKOS/SKOS XL for use with FRBR.
These are only my opinions.
Jeff
From: Jodi Schneider [mailto:jodi.schneider@deri.org]
Sent: Monday, August 09, 2010 6:40 AM
To: Young,Jeff (OR)
Cc: Karen Coyle; public-xg-lld@w3.org
Subject: Re: is FRBR relevant?
Hi, Jeff (& all),
Ok, now I *start* to understand what you're getting at.
Do you imagine an agent using that subject? Among humans, only
catalogers, researchers, and reference librarians are likely to seach
for this subject heading, I think.*
"has as subject" "World War, 1939-1945"
I think what you're saying, though, is "since we've cataloged, wouldn't
it be great to expose the data" -- and that FRBR's "has as subject"
gives a way to do this.
I still haven't figured out why you're asking "is FRBR relevant?" (i.e.
in the subject line).
Maybe your concern is that authority control should give us identifiers
not just uniform headings? I guess Karen's more recent post might be
relevant to this thread:
http://kcoyle.blogspot.com/2009/08/frsad.html
I think you're probably getting at something important, but I'm still
not quite sure what it is.
-Jodi
PS-Any quick intro to suggest for FRSAD? Not up to speed there. I've
added the draft report to my queue:
http://nkos.slis.kent.edu/FRSAR/report090623.pdf
On 7 Aug 2010, at 21:14, Young,Jeff (OR) wrote:
Karen,
Sorry that I raised the issue rhetorically. An explanation would be
better.
The issue is precision and recall
<http://en.wikipedia.org/wiki/Precision_and_recall> of my Google search:
"has as subject" "World War, 1939-1945"
Note that the "has as subject" relationship is straight from FRBR and
"World War, 1939-1945" is straight from LCSH.
My Google search returned a grand total of 2 hits (3 now that Google
indexed this thread). Now imagine a Web-accessible library catalog with
an HTTP URI for each FRBR Work something like this:
http://example.org/work/12345/
Content-negotiation for HTML (the default) could include markup
something like:
<tr>
<th>has as subject</th>
<td>
<a
href="http://example.org/work/?frbr:hasAsSubject=http%3A%2F%2Fid.loc.gov
%2Fauthorities%2Fsh85148273%23concept">World War, 1939-1945</a>
</td>
</tr>
Etc.
The RDF equivalent could be added as RDFa or negotiated from the URI.
Eventually, Google would index these work pages and my search wouldn't
be so disappointing. The same principles apply throughout FRBR.
Jeff
-----Original Message-----
From: Karen Coyle [mailto:kcoyle@kcoyle.net]
Sent: Saturday, August 07, 2010 2:32 PM
To: Young,Jeff (OR)
Cc: public-xg-lld@w3.org
Subject: Re: is FRBR relevant?
Jeff, I don't know what you were expecting when you did this
search,
therefore why you find it to be disappointing. Perhaps you can
explain?
kc
Quoting "Young,Jeff (OR)" <jyoung@oclc.org>:
I've been looking at the relationship between FRBR and
FRSAD over
the
past week.
http://www.ifla.org/functional-requirements-for-bibliographic-records
http://www.ifla.org/node/1297
The fundamental question of FRSAD revolves around the
range on
FRBR's
"has as subject" relationship between Work and other things. One
example
given in the report revolves around the LCSH heading
"World War,
1939-1945", so I typed this query into Google:
"has as subject" "World War, 1939-1945"
Why am I disappointed?
Jeff
---
Jeffrey A. Young
Software Architect
OCLC Research, Mail Code 410
OCLC Online Computer Library Center, Inc.
6565 Kilgour Place
Dublin, OH 43017-3395
www.oclc.org <http://www.oclc.org>
Voice: 614-764-4342
Voice: 800-848-5878, ext. 4342
Fax: 614-718-7477
Email: jyoung@oclc.org <mailto:jyoung@oclc.org>
--
Karen Coyle
kcoyle@kcoyle.net http://kcoyle.net
ph: 1-510-540-7596
m: 1-510-435-8234
skype: kcoylenet
Received on Monday, 9 August 2010 16:39:49 UTC