[open-bibliography] FRBR examples

[Karen Coyle]

On 5/27/10 7:55 AM, Weinheimer Jim wrote:
>
> On the Open Library page for Moby Dick that you gave, do people really have such a need for this summary? And in fact, what is the purpose of the separate page? What if the "works" or "expres
> sions" were only queries that bring the relevant records together for the user to man
> ipulate further?
That is exactly what the OL page does. The page is derived from the data 
in the database. Whether or not it shows summaries or first lines is 
really up to the implementation, but it is essentially what you say: a 
query that brings the relevant records together. At that point, users 
can move in various directions. If you click on the subject
    http://openlibrary.org/subjects/ship_captains_in_fiction
from that page you go to a subject page that has a publishing timeline, 
and a lot of choices for next steps for the user. I think this should be 
the point of the library catalog (or any catalog for that matter), which 
is to help people discover things they might not have originally thought 
about.

This also illustrates something that I talk about a lot: pulling 
together data from the records and making it visual. How long would it 
take a user who gets a list of hundreds of bib records to derive 
something like that timeline? Or to make the links from Moby Dick to all 
of the possible relevant subjects? This is what our systems should be 
doing for users, mining retrieved sets for information.

> For the headings there, Lucene and Zebra indexing can now extract te headings in a much more useful way, e.g. see the Koha catalog at Antioch http://tinyurl.com/2ukcwny and how Zebra indexing extracts the headings into the left-hand column.
>    
Because the data is there, that type of display could be added to OL -- 
it's all a matter of choices, and OL will undoubtedly change over time. 
The main thing is to have the data elements in a usable form, and to 
experiment with different displays and different information 
visualizations.

kc



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