W3C home > Mailing lists > Public > public-lld@w3.org > August 2010

Re: Open Library and RDF

From: <gordon@gordondunsire.com>
Date: Sun, 15 Aug 2010 22:21:51 +0100 (BST)
To: Thomas Baker <tbaker@tbaker.de>, Karen Coyle <kcoyle@kcoyle.net>
Cc: "Young,Jeff (OR)" <jyoung@oclc.org>, public-lld@w3.org
Message-ID: <671850206.939558.1281907311561.JavaMail.open-xchange@oxltgw00.schlund.de>
Tom and others
That is, indeed, the point (ignoring the bracketed text ;-)
The experience of the professional library communities represented by the FR and
RDA committees over the past 150 years indicates that a strong approach is
necessary to meet the needs of their users. Despite powerful economic drivers,
weaker models have generally not been successful in academic and national
library environments (or in larger public reference libraries).
My own experience in discussing FR and RDA in European communities is that there
is little dissent about the need for a strong, controlled approach; most
argument has been about relatively esoteric detail such as the use of
abbreviations. There has been extensive consulation with specialist communities
in Europe, and with the general library communities elsewhere world-wide.
If non-library communities wish to take advantage of the vast experience of
professional librarians represented in the FR and RDA models, then AAA, surely?
Of course, if library communities were (generally) resourced by commercial
organisations, non-heritage government agencies and the like, then I'm sure they
would be as responsive to the needs of such communities as they are to their
actual users.
As it is, both the FR and RDA committees are actively considering developing
unbounded super-class/property versions of their models, resources permitting.
Expect some announcements soon.

On 15 August 2010 at 21:55 Thomas Baker <tbaker@tbaker.de> wrote:

> On Sat, Aug 14, 2010 at 03:19:36PM -0700, Karen Coyle wrote:
> > >Is everyone involved in the process happy with this?
> >
> > This depends on what you mean by "involved in the process"? The FR 
> > committees and JSC (developers of RDA) are in accord on this 
> > principle, but there is considerable dissent in the US library 
> > community, in particular from specialist communities who tend to have 
> > different definitions of what constitutes a W,E,M. These differences 
> > are simmering in the background because as yet there is no 
> > implementation of FRBR as a data carrier. If this "strong" view of 
> > WEMI is constrained in the carrier, some folks are not going to be 
> > able to create metadata that expresses their community view.
> Or as Halpin and Hayes put it [1]:
>     Just because a construct in a knowledge representation
>     language is prescribed a behavior using formal semantics does
>     not necessarily mean that people will follow those semantics
>     when actually using that language "in the wild".  This can
>     be laid down to a wide variety of reasons.  In particular,
>     the language may not provide the facilities needed by people
>     as they actually try to encode knowledge, so they may use a
>     construct that _seems_ close enough to their desired one.
>     A combination of not reading specifications -- especially
>     formal semantics, which even most software developers and
>     engineers lack training in -- and the labeling of constructs
>     with "English-like" mnemonics naturally will lead to the use of
>     a knowledge representation language by actual users that varies
>     from what its designers intended.  In decentralized systems
>     like the Semantic Web, this problem is naturally exacerbated.
> This suggests that it might be wiser, especially with a
> relatively untested language such as FRBR, to follow the
> principle of minimal ontological commitment [2]:
>     An ontology should require the minimal ontological commitment
>     sufficient to support the intended knowledge sharing
>     activities. An ontology should make as few claims as possible
>     about the world being modeled, allowing the parties committed
>     to the ontology freedom to specialize and instantiate the
>     ontology as needed. Since ontological commitment is based on
>     consistent use of vocabulary, ontological commitment can be
>     minimized by specifying the weakest theory (allowing the most
>     models) and defining only those terms that are essential to
>     the communication of knowledge consistent with that theory.
> I take you to be saying that the official "strong" view
> reflects a specialized view of one part of the community.
> Do the FR and RDA committees anticipate that the vocabularies
> will only be used by (and of interest to) bibliographic
> experts with MLS degrees?  Do they not anticipate that
> commercial companies, government agencies, or other types of
> organizations might want to embrace FRBR concepts and adapt
> them to their needs?  Raising the bar ontologically would
> help ensure that their use is limited to controlled contexts
> (or at any rate used incorrectly outside those contexts).
> Or is that perhaps the point?
> Tom
> [1] http://events.linkeddata.org/ldow2010/papers/ldow2010_paper09.pdf
> [2] Gruber, Thomas. 1995. Toward Principles for the Design of
>     Ontologies Used for Knowledge Sharing. International
>     Journal Human-Computer Studies 43(5-6): 907-928.
> --
> Thomas Baker <tbaker@tbaker.de>
Received on Sunday, 15 August 2010 21:22:31 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 16:18:57 UTC