From: "Rupp-Serrano, Karen J." <[log in to unmask]>
Date: Fri, 21 Mar 2014 20:06:51 +0000

Colleagues,

With this message I am providing a summary of responses received to my
query about licensing language for text and data mining.

One respondent indicated the following language has been proposed for
use by their institution, but so far can report no results on its use:

Data/Text Mining.  Licensor shall provide either online at its web
site, through a third party service (such as LOCKSS or Portico), or to
Licensee in mutually agreed physical media, one full copy of the
Licensed Materials in raw data format.  The raw data may be used by
Licensee and Authorized Users to perform text, image, and/or data
mining functions and algorithms for legitimate academic research and
other educational purposes in accordance with the terms of this
Agreement.

A second respondent noted language from CDL:

Text Mining.  Authorized users may use the licensed material to
perform and engage in text mining/data mining activities for
legitimate academic research and other educational purposes.

This same respondent brought up Peter Murray-Rust's argument against a
200 character limit and the limitation to use mined data for
non-commercial purposes.  The respondent also questioned the CC BY-NC
license, arguing that much extracted data is inherently not
copyrightable content, and noted that "content mining" may be a better
phrase because it allows for the extraction of images, sound, etc.

Ann Okerson noted that draft language on this issue is currently being
considered as part of the revision of the LIBLICENSE Model License.

Thanks to those of you who so kindly responded to my query.  I have
responded to the publisher in question but have, as of yet, received
no reply.

Karen Rupp-Serrano
University of Oklahoma Libraries
[log in to unmask]


From: "Rupp-Serrano, Karen J." <[log in to unmask]>
Date: Fri, 28 Feb 2014 18:01:39 +0000

Colleagues,

On behalf of a member of my campus community, I recently contacted a
publisher who is offering an API key for text mining.  The publisher
responded with an amendment to our current subscription agreement
which enables all users to

"access the text and data mining service online via an API to
continuously and automatically extract and index and/or process
information from the subscribed products, and load and integrate the
results on the subscriber's text-mining system for access and use by
authorized users"

And

"distribute the output externally, which may include a few lines of
query-dependent text of individual full text articles or book chapters
which will be up to a maximum length of 200 characters surrounding and
including the text entity matched ('snippets') or bibliographic
metadata.  Where snippets and/or bibliographic metadata are
distributed, they should be accompanied by a DOI link that points back
to the individual full text article or book chapter.  Further, the
output should include a creative commons proprietary notice "(C) Some
rights reserved.  This work is distributed under the terms of the
CC-BY-NC 4.0Š"  Text and data mining svcs online may be accessed by
vendors or other third parties retained by the subscriber only with
the express permission of publisher and for the index and/or process
information purposes of the subscriber"


I would be most interested in hearing from any of you as to what you
would suggest (or have successfully negotiated) be modified, struck,
added, etc., in this language to best serve the interests of the
institution and its researchers.  Perhaps we can begin to develop some
standard language that all of us can use as we move forward in this
emerging area.

Please respond directly to me (contact below) and I will summarize the
findings anonymously for the list.

Thank you,

Karen Rupp-Serrano
University of Oklahoma Libraries
[log in to unmask]