Why is there a need to post a list of OA articles?

Eric, I agree that access/rights metadata is very important,
and 
is not as easily available as it should be.

One source of access metadata that can be useful is that
which is 
provided by publishers (including some optional OA
publishers) to 
PubMed.

This metadata allows you to restrict a search in PubMed to
only 
include articles where PubMed has a link to a free fulltext.

e.g. a search for:

pnas [ta] AND free full text[sb]

finds all 83190 PNAS articles that have free fulltexts
(including 
both the latest optional OA articles, and the back archive)

Interestingly, because of PNAS's liberal policy on making
even 
the subscriber only content available free after a
relatively 
short embargo, the total number of PNAS articles that are 
subscriber only is small at any one time:

pnas [ta] NOT free full text[sb]

finds only 1545 articles

Matthew Cockerill, Ph.D.
Publisher
BioMed Central ( http://www.biomedcentra
l.com/ )
Email: mattbiomedcentral.com

> -----Original Message-----of
> [mailto:owner-liblicense-llists.yale.edu]On Behalf Of
Eric Hellman
> Sent: 19 April 2006 01:26
> To: Martin Frank; liblicense-llists.yale.edu
> Subject: Why is there a need to post a list of OA
articles?
>
> Libraries need to know what items are available to
their users. 
> Automated systems are playing an increasing role in
matching 
> readers to articles.
>
> In an increasingly typical scenario, a reader discovers
an 
> article of interest in a search engine- Google Scholar,
Pubmed, 
> Scopus, etc. The search engine have an interaction with
an 
> institutional link server to display to the user
whether or not 
> full text for the item can be found in their library.
Google 
> Scholar does this by uploading holdings files from the
library; 
> Scopus does this dynamically using an
"image-based linking" 
> technique.
>
> Today, link servers in libraries that do not subscribe
to your 
> journal will tell their users that the OA article is
not 
> available in their library, because there is no way to 
> determine automatically that the article is in fact OA.

> Potential readers will miss the article despite the
fact that 
> the author has paid an OA fee.
>
> It's something to work on- I would love to hear from
publishers 
> that would like to see this type of access metadata
more easily 
> available to libraries.