Open Access News

News from the open access movement

Saturday, June 02, 2007

Asking a society journal to convert to OA

The researchers planning an OA journal on computational linguistics and natural language processing have now submitted an official proposal to the executive committee of the Association for Computational Linguistics.  It's undated but was apparently sent on June 1.  (Thanks to Hal Daume III.)  Excerpt:

We propose that the ACL journal Computational Linguistics convert its publishing model from subscription-based to open access. Open access journals are provided freely to all, eliminating access barriers to our scholarly writings and making uniform their availability, consistent with the purpose of the Association to “promote research and development activities in the field of computational linguistics” and “provide information on computational linguistics to the general public”. Open access provides wider sharing of knowledge and the acceleration of research, and is thus in the best interest of authors, potential readers, and the computational linguistics community as a whole.

Concomitant with this change would be the following changes to the publication process:

  • The process would take advantage of advances in journal management software systems, enabling web-based workflow management of the submission, editor assignment, reviewing, and issue processes.
  • Author permission would be changed to a Creative Commons license, so as to enable free accessibility of the articles for all purposes, including data mining and use as an experimental corpus.
  • Authors would reclaim responsibility for typesetting, eliminating the anachronisms and problems of the galley proof change process.
  • Articles would be issued upon acceptance of the final typeset version, so that the journal would carry no backlog. This would dramatically shorten publication delays.
  • Articles would be accepted independently of artificial page limitations. This would enable, though not require, accepting more articles for publication if the supply of high-quality articles allowed.
  • A print edition would be published according to the principles of open-access print archiving, thereby maintaining the true archival aspect of the journal. Microtome publishing has already agreed to serve as the print-archive publisher, though MIT Press may also be willing to perform this task.

In summary, the model and process of publication would be essentially identical to that used by the Journal of Machine Learning Research, the highest impact factor journal in artificial intelligence and the second highest of any computer science journal....

The proposal also contains a useful FAQ for any journal considering the switch. 

PS:  Kudos to the authors, Hal Daume III, Kevin Duh, Ryan McDonald, Fernando Pereira, and Stuart M. Shieber, for writing an exemplary proposal.  For background on their efforts to launch an OA journal, see my posts from May 16 and May 22, 2007.