Retraction Watch, a blog about retractions in academic journals, published on Monday an interesting post about Springer and IEEE retracting over 120 computer generated articles. This post was based on an article published in Nature. A researcher, Cyril Labbé, has for a couple of years cataloguged computer genereted articles from 2008 to 2013. These articles were published in Springer and IEEE’s conference proceedings. Many of the articles have made-up author-names but it seems that even legitimate researchers are connected to these articles, although they do not have anything to do with the article or even have been aware of the fact that these articles exist.
Labbé has developed an automatic way to discover articles generated with SCIgen – software which combines strings of words to create papers in computer science. The software was created in 2005 by MIT in order to show that conferences were accepting papers which were not properly reviewed.
The interesting part in this story is that it hits Springer and IEEE, two traditional scientific publishers. Last fall, Science published an article by journalist John Bohannon pretending to be a journalist. He wanted to show that open access publishers would publish anything. The study shows that Bohannon’s gibberish article was accepted by approx. 150 open access journals of approx. 300 open access journals. The study was widely criticized because none of the traditional subscription based articles were included in the selection.
Both studies show that the peer review system is broken or misused. One of the readers of Retraction watch left a comment asking “the fraud is so wide and so broad, will retractions even be a solution to fix this mess?” Maybe one way to work against this problem is to continue to develop initiatives such as open peer review, British Medical Center is working with it (check the example how open review looks), PubMed has started with PubMed Commons where articles are commented after publication and Peer J recommends authors to be open with the review process. Perhaps open peer review is not the definitive solution to the problem but it is one way of avoiding or fighting it.