Thousands boycott new Nature journal about machine learning

More than two thousand researchers have signed a petition to boycott a new Nature journal over the fact it will be available only by subscription.

The new journal — Nature Machine Intelligence, about machine learning — plans to charge readers for access, unlike most other journals in the field. The researchers who signed the petition have pledged not to submit their work to the new journal, and will decline to review or edit papers for it, as well.

Most journals published by Nature Publishing Group are available only by subscription — but that doesn’t work for the machine learning community, the signatories argue:

[We] believe the adoption of this new journal as an outlet of record for the machine learning community would be a retrograde step. In contrast, we would welcome new zero-cost open access journals and conferences in artificial intelligence and machine learning.

The petition was released on Saturday around 2000 UTC time, according to coordinator Tom Dietterich of Oregon State University; by Tuesday morning, it had 2160 signatures. Dietterich told Retraction Watch:

I expected this initiative to be strongly supported by the research community, but I am surprised at how quickly the signatures have accumulated.

He added:

Free and open access to knowledge is important in all fields. It is particularly important for students and faculty whose universities cannot afford the subscription fees for closed-access journals or cannot afford to pay (as authors) for their papers to be open access. Open access speeds up scientific progress by enabling anyone anywhere on earth to read the latest research and make their own contributions.

The machine learning community has embraced openness in publishing papers, software, and data sets. We have previously fought with Kluwer Academic Press…over open access to the journal Machine Learning (which I served as Executive Director for 6 years from 1992-1998). This resulted in a partial victory in that papers in MLJ are now open access 1 year after publication. It also resulted in the creation by Leslie Kaelbling (MIT) of the Journal of Machine Learning Research, which continues to be the most prestigious place to publish journal-length machine learning papers.

According to the petition, in 2001, the editorial board resigned from a subscription-based journal, Machine Learning, now published by Springer (which has merged with the publisher that includes Nature Publishing Group), and formed the open-access journal the Journal of Machine Learning Research (JMLR).

In response to Dietterich’s tweet about the petition, Nature Machine Intelligence posted over the weekend:

A spokesperson for Nature Machine Intelligence told us:

We support the machine learning community’s view that open access and preprint servers, such as arXiv, have important roles in the dissemination of research. At Nature Research we are also achieving this by helping the community to freely share their discoveries by encouraging preprint posting and data- and code-sharing. We continue to extend access to all Nature journals in various ways, including our free SharedIt content-sharing initiative, which provides authors and subscribers with shareable links to view-only versions of published papers.

Selective journals like Nature Machine Intelligence — which involve substantial editorial development, aim to provide high levels of author service and publish informative, accessible content beyond primary research — require investment. At present, we believe that the fairest way of producing these journals, which ensures their long-term sustainability as a resource for the widest possible community, is to spread these costs among many readers — instead of having them borne by a few authors.

Dietterich told us:

We hope to prevent the establishment of an expensive closed-access journal in the machine learning field. This could unfold in a number of different ways. For example, Nature Publishing Group could make their new journal free and open access or they could decide that there is not a viable business in the machine learning area.

The new journal is scheduled to launch in January 2019.

Like Retraction Watch? You can make a tax-deductible contribution to support our growth, follow us on Twitter, like us on Facebook, add us to your RSS reader, sign up for an email every time there’s a new post (look for the “follow” button at the lower right part of your screen), or subscribe to our daily digest. If you find a retraction that’s not in our database, you can let us know here. For comments or feedback, email us at team@retractionwatch.com.

15 thoughts on “Thousands boycott new Nature journal about machine learning”

    1. It’s not a new idea in this field: the Journal of Artificial Intelligence Research (JAIR) is zero cost open access and has been around since 1993.

      1. Yah, I was only an AI grad student for two years (picked very much the wrong program, but I was naive), and back in the early ’90s, everything happened at conferences and in the “proceedings” thereof.

  1. Open access is the best thing ever happened in machine learning. That is one of the major reasons that the field is prospering. If you take away that only thing left will be hype.

  2. The big hole in the idealism of open access (which in principle is wonderful) is that somebody has to pay for the overall organization, servers, etc.
    In these days of hacked budgets it seems to be yet another cost dumped on PIs. The cost of complying with NIH open access rules for my last publication was equivalent to the price of two microfuges.

  3. Recently Nature Publishing Group has been creating several field journals, which they try to market as the most selective and highest impact in their respective fields. I find their attempt to hi-jack fields with established (often open-access) journals very annoying and I will not submit to their journals or review for them.

    Although scientists generally don’t think that “Nature something” means good quality, unfortunately department heads and deans are often suckers for anything Nature or Science.

  4. Mmm let’s make open access journals without pushing the authors to pay the fees ! the only way is by PrePrints ! and Arxiv is doing a great job in this perspective !
    The other traditional way (pay to publish) for openness and free science etc Bla Bla Bla is just a stupid idea ! Why should I pay to publish my papers ?

    1. Because then your papers will be peer-reviewed, potential errors spotted and corrected, and it’ll make a more solid and more readable contribution to science that other people will be more inclined to trust (with a pinch of salt, as usual). This added value, as well as final hosting for decades, has a cost.
      I’m not saying journals have you pay a fair price for publishing, far from it, but it can’t be free as in “free lunch”.

  5. It is worth pointing out some more machine learning heavyweights that have signed the petition:

    * Geoffrey Hinton, Google
    * Ian Goodfellow, Google Brain
    * Yoshua Bengio, U Montreal
    * Yann LeCun, NYU & Facebook

    They do know about the role of arXiv and open-access for their research.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.