From: Ivy Anderson <[log in to unmask]>
Date: Sun, 28 Oct 2012 23:44:33 +0000

Thanks to Ann for highlighting the work of the UC Santa Cruz
Genocoding Project.  Libraries can help with this.  The California
Digital Library (CDL) negotiates systemwide agreements for the UC
system, and we have worked hard to include text mining provisions in
our licenses for a number of years.  However, researchers are often
unaware of the library's efforts in this regard and may not think to
reach out to their libraries for assistance.  In many cases, we may
have already negotiated the rights they are seeking.  How those rights
are operationalized is another matter, and one in which libraries
should also consider playing a more active role.

We in the library community can do more to connect with our
researchers on these issues and invite them to enlist our help in
working with publishers and other information providers on text mining
arrangements.


- Ivy

Ivy Anderson
Director of Collections
California Digital Library
University of California, Office of the President
[log in to unmask]
http://cdlib.org

-----Original Message-----

From: Ann Shumelda Okerson <[log in to unmask]>
Date: Sun, 28 Oct 2012 04:17:38 -0400

1.  Text mining (UCSC Gencoding Project).  Interesting information
here -- I was happily surprised to see that already there is a lot of
text mining opportunity, with growing improvement in coverage, though
a long way to go.  Also pleased to see a project that systematically
contacts publishers and reports results of those contacts.  Are there
other systematic inquiries and studies for research in other fields?
Do any of our liblicense-l readers know?

This information came courtesy of Kathleen Shearer, Canadian
Association of Research Libraries

Dear Ann:  Here is the url for the text mining permissions of
publishers as identified by UCSC Genocoding Project:

http://text.soe.ucsc.edu/progress.html

* Current coverage of Pubmed 30% (2.7 million articles)

* We have already indexed 8 million documents from Elsevier and
PubmedCentral. They represent roughly 30% of PubMed (all years,
~1940-today). PubmedCentral includes text from most open-access
publishers.

Requests for permission sent to publishers

* Out of the 6 million articles published since 2000 in Pubmed, we
concentrate on publishers with more than 1000 articles.

* There are 527 publishers that fulfill these criteria.

* We have contacted the following ones. As shown, not all publishers
require permission for text mining.

Web crawling has started in June 2012,


*******

2.  Charleston Conference Session:

Concurrent Session: Text Mining Rights from Three Perspectives
Thursday, November 8
3:15 PM - 4:00 PM
Francis Marion Hotel, Carolina Ballroom

Speakers: Teresa Lee (E-Resources & Access Librarian, University of
British Columbia), Heather Piwowar (Postdoc, Duke and University of
British Columbia) and Judson Dunham (Senior Product Manager, Elsevier)

*******

Ann Okerson/CRL