Why are thousands of AI researchers boycotting the publication of papers in academic journals such as Nature?

byFranck V.

Nature(Nature) and other academic journals publish various academic papers every year but research papers on machinery learning attracting attention as AI related research fields are open to anyone open free access journals etc. that can be used free of charge There are many things. Neil Lawrence, a professor of machine learning at the University of Sheffield, explained about why there are many cases where it is not open to scientific journals and why it is necessary to subscribe for a fee but to be open to anyone who can read it for free at anyone. I will.

Why thousands of AI researchers are boycotting the new Nature journal | Science | The Guardian
https://www.theguardian.com/science/blog/2018/may/29/why-thousands-of-ai-researchers-are-boycotting-the-new-nature-journal

More than 200 years ago, astronomersGiuseppe PiazziIn historyDiscover asteroids for the first timeDid. Piazzi writes a letter to his friend Franz von Zac and tells the discovery of the asteroid. After that, Zack duplicated and sent Piazzi letters to all astronomers in Europe, and conveyed the existence of asteroids to researchers around the world, but then there was a way to convey information to the world in the moment like the Internet at that time Because it did not, it took a long time to spread the existence of Ceres. In addition, it seems that there was also a problem that since the asteroid had already disappeared in the sun's light at the time when the discovery of Piazzi spread, the fact can not be confirmed.

It was the saving of such a crisis of Piazzi that the German astronomerCarl Friedrich Gaussis. Gauss predicted Ceres 's orbit by applying Kepler' s law to planetary motion, and helped rediscover Ceres. From this incident Piazzi and Gauss are deepening friendship and they seem to have shared each other's learning freely. And he seems to have felt the need to pay the result for "the role of disseminating the dissertation" which Zack played, and this is a general academic term "to publish private papers to part of pay readers" Mr. Lawrence wrote that it led to the beginning of the model adopted in the magazine.

byKyler Boone

"Modern data" was the observation data of the planet for Gauss and others, but in the field of machine learning it is medical images, audio languages, Internet data etc. By using this "modern data", advanced diagnostic systems, speech recognition systems, advanced machine vision of automatic driving cars, etc. are being developed, and machine learning is like a spine supporting AI development It is Mr. Lawrence. In other words, Mr. Lawrence asserts that the paper on machine learning is indispensable for developing state-of-the-art technology.

In researchers who are active in such machine learning field, there are few people who find the necessity of posting research results in closed publication models, that is, general scientific journals. On the contrary, in modern times it is freely available via the InternetOpen Access JournalThere are also many researchers who publish their research results free of charge, so even those who feel that the form "check papers at a fee" used in common academic journals is outdated It is said that there are many.

byAlex Holyoake

Research funding for machine learning is often paid by governments, etc. In other words, it means that the tax paid by citizens is research funds. If tax is research funds, it is a strange structure to pay money to read the results, Mr. Lawrence. In addition, the field of machine learning is "a few fields demonstrating that academic fields can survive and succeed without the involvement of commercial publishers".

However, it is not so much whether publishers are hindering entry into the market "machinery learning thesis publishing platform", but "Nature Machine Intelligence" etc. which publish research papers in the machine learning field etc. already exist .

Nature Machine Intelligence


In the machinery learning field with a feeling that an open platform for sharing ideas has already been established, the advent of paid academic journals such as Nature Machine Intelligence is not necessarily a matter of concern. Paid scholarly journals and others are sometimes used to secure the quality of research papers, for example, publishing a paper in a famous academic journal is a great help if you want to widely appeal your research results. However, the academic journal is the dealer of the article to the last, researcher who is writing the paper. The quality will also depend on the researcher, Mr. Lawrence points out. That is, if quality is accompanied it claims that there is no reason to publish the paper in a fee-based academic journal or the like.

In addition, Mr. Lawrence says, "Many of our research communities believe that brands such as Nature are being used as collateral in poorly academic quality," said a paid academic journal on machines We are seeing the current situation eroding the learning field with a negative view.

These ideas are not limited to Mr. Lawrence, but more than 3,000 researchers including influential researchers post papers on AI and machine learning in a paid academic journal "Nature Machine Intelligence" I am issuing a statement that refuses to do.

Home | Statement on Nature Machine Intelligence | Oregon State University