March 2, 2014
Jargon Or Gibberish? How Springer And Other Scientific Journals Were Fooled By Computer-Generated Papers
redOrbit Staff & Wire Reports - Your Universe Online
For the average reader, the line between jargon-heavy scientific research and unintelligible gibberish is a fine one, but apparently Joe Sixpack isn’t the only one who occasionally has trouble telling the difference.
On Thursday, scientific journal publisher Springer announced that it would be removing 16 fake research papers from its archives after learning that they were essentially computer-generated nonsense. The firm said that they were tipped off by Dr. Cyril Labbé, a French researcher who published research on how to detect computer-generated papers last January in the journal Scientometrics.
According to AFP reporters Richard Ingham and Laurent Banguet, the fraudulent papers were created using SCIgen, a free program used to create pseudo-academic research. They were then submitted to computer science and engineering conferences and then printed in specialized, subscription-only publications.
Ingham and Banguet said that SCIgen allows users to produce “impressive-looking” fake research studies “stuffed with randomly-selected computer and engineering terms.” The document comes “complete with fake graphs and citations – essential features in scientific publishing” and “includes recent references to famous scientists.”
The AFP reporters also included an example of the SCIgen content: “Constant-time technology and access points have garnered great interest from both futurists and physicists in the last several years. After years of extensive research into superpages, we confirm the appropriate unification of 128-bit architectures and checksums.”
Similarly, on Friday, Jemima Lewis of The Telegraph detailed how she was able to use the program to develop a fake academic paper entitled “The Impact of Amphibious Models on Interposable Algorithms,” which featured what she referred to as a “memorable” opening line: “The exploration of lambda calculus has investigated neural networks, and current trends suggest that the construction of Byzantine fault tolerance will soon emerge.”
“We are in the process of taking the papers down as quickly as possible. This means that they will be removed, not retracted, since they are all nonsense. A placeholder notice will be put up once the papers have been removed,” Springer said in its statement. “Furthermore, we are looking into our procedures to find the weakness that could allow something like this to happen, and we will adapt our processes to ensure it does not happen again.”
“For the moment, we are using detection programs and manpower to sift through our publications to determine if there are more SCIgen papers. We have also reached out to Dr. Labbé for advice and collaboration on how to go about this in the most effective manner,” it added. “We are confident that, for the vast majority of the materials we publish, our processes work. When flaws are detected by us, or brought to our attention by members of the scientific community, we aim to correct them transparently and as quickly as possible.”
Springer’s disclosure highlights a growing problem in the scientific publishing industry. Dr. Labbé’s work uncovered a total of 120 computer-generated papers that had been published by respected institutions in the US, Germany and China, Lewis explained. Over 100 of those papers were published in by the US Institute of Electrical and Electronic Engineers (IEEE), The Guardian noted.
SCIgen was originally developed by MIT graduate students Jeremy Stribling, Dan Aguayo and Maxwell Krohn in 2005. The trio wanted to expose how scientific contests would accept any type of academic papers so long as the authors were willing to pay the registration fees, the Guardian said.
The program took just a few days to complete, and the nonsensical paper it produced was sent to and accepted by a conference. While their efforts “revealed a farce that lay at the heart of science,” the UK newspaper said that SCIgen is “the hoax that keeps on giving” because Stribling, Aguayo and Krohn made the software free to download and, as Dr. Labbé’s research proves, “scientists have been using it in their droves.”
“There will always be individuals who try to undermine existing processes in order to prove a point or to benefit personally. Unfortunately, scientific publishing is not immune to fraud and mistakes, either,” Springer concluded. “The peer review system is the best system we have so far and this incident will lead to additional measures on the part of Springer to strengthen it.”