Re: Incubate best practices

I like Karen's suggestion.

The library community copes with a growing array of metadata standards and
data formats.  It is perhaps too much to agree on a common model for all
library data, but agreeing on such a model for some of the broadly shared
entities will allow the community to interoperate more effectively, and to
make library data more prominent and relevant on the web.

Achieving such a model without being swallowed by the enormity of the task
is a daunting challenge, and likely larger than this group can manage.
Perhaps an agreement on a set of common linked data patterns, and division
of the problem into tractable chunks, with clean, simple interfaces between
the chunks?

An important deliverable of the LLD XG could be the elucidation of the
characteristics of a common model that embraces the legacy of MARC data and
the abstractions of FRBR, and makes evident how such a model could be grown
coherently to include the varieties of other metadata relevant to the
community?

I suspect there is a group of potential collaborators among us who might
want to go further with this than the group as a whole.

On 6/15/10 10:17 AM, "Karen Coyle" <kcoyle@kcoyle.net> wrote:

> LLD'rs
> 
> Since library data has some common patterns, it seems like one project
> to incubate would be the creation of examples of how to present those
> patterns in linked data, and some best practices for "linked data
> friendly" output from traditional library data. This would save
> everyone a lot of time in developing linked data for libraries, and
> would probably have the result of bringing more consistency overall.
> 
> As an example, there could be a best practice for the description of
> personal names and name variants, a topic that has come up recently on
> the foaf list. This is a pretty complex problem, and it would be a
> shame to have to solve it multiple times in different projects, with
> different results.
> 
> Best practices aren't mandatory, so no one HAS to follow them; they
> exist to guide those who desire guidance, and to save people time.
> They are not usually as formal as standards, and could be entirely in
> the form of examples and short explanations. There can be competing
> best practices, so that one has a choice should there be more than one
> "good way" to solve a problem.
> 
> kc

Received on Tuesday, 15 June 2010 18:07:51 UTC