Academic Paper Spam

There seems to be an epidemic of computer-generated nonsense academic papers.

Labbé does not know why the papers were submitted -- or even if the authors were aware of them. Most of the conferences took place in China, and most of the fake papers have authors with Chinese affiliations. Labbé has emailed editors and authors named in many of the papers and related conferences but received scant replies; one editor said that he did not work as a program chair at a particular conference, even though he was named as doing so, and another author claimed his paper was submitted on purpose to test out a conference, but did not respond on follow-up. Nature has not heard anything from a few enquiries.

In this arms race between fake-paper-generator and fake-paper-detector, the advantage goes to the detector.

Posted on March 7, 2014 at 6:13 AM • 28 Comments

Comments

Bob S.March 7, 2014 6:56 AM

Apparently no real people actually read the papers. I've suspected that for a long time.

There is a cautionary moral to this story, somewhere.

JeffMarch 7, 2014 7:15 AM

Bob, I suspect there are tiers of academic conferences and publications. The higher, more prestigious tiers are all human-reviewed content. The lower ones exist to be crammed full of stuff since it is publish or die for many academics.

ElizabethMarch 7, 2014 7:22 AM

I read something a while back about how many university IT programs in... India, I think, required students to have a paper accepted by an international journal as a requirement for graduation. And so many fake papers, and fake conferences to accept them, made their way to the web. My guess is that this is a similar situation. Of course none of the authors are going to fess up if their only goal was to get published.

When I was doing research for an academic paper in college I came across a few papers from Asia that were written in English but completely incomprehensible. I wondered how any editor could possibly let them through. So I'm saddened but not surprised that many published papers are never actually read much less properly reviewed.

Nicholas WeaverMarch 7, 2014 7:23 AM

THe reason for so much spam is the rise of bogus conferences. A decade ago it was just the whole cybernetics crap in Florida, which was once a year. Now as an academic I swear I get a CFP-spam a day.

Whipping off a SCIgen paper is a great way to check just how bogus they are, and I've (vaguely) contemplated making an auto-garbage submit site.

Although the best response to a spam conference has to be
Get me off Your Fucking Mailing List" by
Mazieres and Kohler.

Clive RobinsonMarch 7, 2014 8:45 AM

@ Jeff, Elizabeth,

Yes the so called "international conferences" in India are a compleate scam, a student has to submit one or two papers, but they also have to present them as well. However they have to pay "full conferance fees" a publishing fee and also travel accommadation and subsistance, whch I'm told can be equal in total to a years income for a "middle class" earner.


The only thing we don't know for certain is if there is any kick back from the confrences to the course asministrators.

I'm sure as in the past we will get a barage of Indian people complaining, but to be quite honest I realy think the Indian Government should get a grip on the situation because it does neither India or their students any favours, in fact I know it's made several employers in the UK insist on proof of non-indian qualifications with CVs, not because they want to be racist, they just don't want to fall foul of "bought qualification" rules in imigration legislation, which could result in significant fines...

Sandy HarrisMarch 7, 2014 10:03 AM

I worked for a few years as an editor in the Comp Sci dep't at Shanghai Jiaotong U (one of China's top schools), improving the English in journal & conference papers mostly by grad students. Getting two papers published was a requirement for the PhD.

Most stuff I saw was pretty good technically. As one might expect, Lai's students were doing some interesting crypto research and lots of other stuff was quite good.

There was some rubbish, though, and lots of attempts to get the required two papers out of one good idea. It might be fairly easy for journals to detect some of those; just look for previous papers by the same authors. If the title is related & the new paper does not cite the previous one, that is a dead giveaway.

RonMarch 7, 2014 10:14 AM

"In this arms race ... the advantage goes to the detector" -- this is because all the fake papers in the study are created by the same software (SCIgen), which is not even under active development. So not much of an arms race.

lax-goalieMarch 7, 2014 10:55 AM

In this arms race between fake-paper-generator and fake-paper-detector, the advantage goes to the detector. For now.

There. Fixed that for you.

RonnieMarch 7, 2014 11:56 AM

If an infinite number of fake paper generators generate an infinite number of fake papers, will one of the papers be the Unified Field Theory/P versus NP/[insert favorite mystery here]?

WitnessMarch 7, 2014 12:11 PM

Academic fraud is hardly limited to students. I recently read that one half of all published research (all disciplines) cannot be reproduced. There is often more than errors in methodology at work here.

When my brother was doing post doc at an Ivy League Medical School (Founded By Benjamin Franklin!). The lead researcher would sneak into the lab late at night, and erase and alter the notes of grad students (kept in pencil!). They caught on, and proved she was doing it. The administration did nothing . She published a paper claiming that 30% of all victims of a horrific neurological disorder (always fatal), carried a certain gene. Her "discovery" even made the NYT! It took more than a decade for two other studies to refute this.

A few months later, she was quietly asked to leave. She had ascertained her access to NIH grants, the med school was in a similar position. What's a few million and human suffering (as researchers barked up the wrong tree)?

I laugh when academics accuse industry funded scientists of "bias". Pot - meet kettle.

PS: Great movie - Krippendorf's Tribe

JonMarch 7, 2014 3:18 PM

When these computer-generated nonsense papers first came out, there was a big todo about how one such paper was accepted by a conference. Just for yuks, I took a look at the conference announcement. It looked to me like complete nonsense too. I started wondering if maybe there was a program generating nonsense conference announcements too, and maybe what we were seeing was dueling nonsense generating programs - one generating nonsense papers and one generating nonsense conference announcements.

QMMarch 7, 2014 3:34 PM

Just a bit further down this path, and we'll be talking about Turing Tests, Blade Runner, and Cherry 2000.

FluffytheObeseCatMarch 7, 2014 8:13 PM

@Witness
The existence of fraud in academia does not obliterate the immense power of money. Bias induced by industry funding (and the threat of it being pulled) is a real problem.
Since you argued by anecdote, I'll do the same. My father-in-law was a fairly high calibre Alzheimer's researcher back in the day. His research focused on the impact of bioavailable aluminum on the progress and severity of the disease. He was heavily funded by a major pharmaceutical company...... up until it acquired a well known brand of Al-rich antacid. His funding was yanked immediately after the acquisition. It left him scrambling and slowed his research and publication rate down for a while.

He transitioned into a more managerial role at his medical school around that point in time. The move was a step up, but, the bait and switch funding very effectively eradicated a legitimate line of research..... for pecuniary reasons.

DBMarch 7, 2014 9:06 PM

According to Darwin's theory of evolution, we should let these fake papers run wild... There will be tremendous breakthroughs and discoveries in them if we let them go for millions of years... They'll even eventually spontaneously turn into a new life form!

Scott "SFITCS" FergusonMarch 7, 2014 10:11 PM

@DB

According to Darwin's theory of evolution,

You haven't actually read it have you?


we should let these fake papers run wild... There will be tremendous breakthroughs and discoveries in them if we let them go for millions of years... They'll even eventually spontaneously turn into a new life form!

And a large number of monkeys with typewriters will write better jokes? Then again.... how can they not?

DBMarch 8, 2014 2:51 PM

@Scott

Then again.... how can they not?

Exactly, you got it :)

I blame Ronnie ;)

Clive RobinsonMarch 8, 2014 3:20 PM

@ DB,

    I blaim Ronnie

Which one?

The one that comes immediatly to mind is the one that "played straightman to a monkey" a while before just playing the monkey in US Politics, so much so a UK satirical program had him staring in "The President's brain is missing"...

And another UK saterical program had a "campfire folk singers" sketch --reminicent of those seen on some US C&W shows-- where they sang "I beleive that even pigs and DC10's can fly, but I can't beleive Ronnie Raygun is President"...

OBMarch 10, 2014 9:11 AM

@DB:
"According to Darwin's theory of evolution, we should let these fake papers run wild... There will be tremendous breakthroughs and discoveries in them if we let them go for millions of years... They'll even eventually spontaneously turn into a new life form!"

According to Darwin there also is SELECTION,
which you just ignored.

vwmMarch 10, 2014 1:50 PM

I am not sure if the “advantage goes to the detector.” Labbé uncovered quite a bunch but some of those papers have gone undetected for years. And we have little knowledge how much generated stuff remains unnoticed.

If someone received benefits (e. g. a degree, funding, a position) from those papers, chances are that he or she can keep those:

  • At least in Germany the limitation period for Bachelor- and Master-Degrees is rather short.
  • If someone received funding or a position, it might be hard to maintain afterwards that
    1. he or she submitted the fake stuff themselves
    2. he or she would not have gotten the funding/position without it

(I assume, plagiarism is more of a problem in Germany than generated papers, but that's just a wild guess. And who says that you can not build a generator that duplicates parts of existing publications?)

vwmMarch 10, 2014 1:51 PM

@Ronnie That will not work [1].

By the way: I guess Jorge Borges (like Paul Erdős) would be a great co-author if you can not resist the temptation to submit a fake paper to a fake conference.

[1] Jorge Luis Borges: “La biblioteca de Babel”, 1941

Terry ClothMarch 11, 2014 12:04 PM

@Kent Johnson:

Not to worry. :-)

At least in the case of the study you referenced, there are no traces in the literature—on the occasions the fake was accepted, the submitters withdrew the paper, referring to an “embarrassing mistake”: “a serious flaw in our experiment which invalidates the conclusions.”

Leave a comment

Allowed HTML: <a href="URL"> • <em> <cite> <i> • <strong> <b> • <sub> <sup> • <ul> <ol> <li> • <blockquote> <pre>

Photo of Bruce Schneier by Per Ervland.

Schneier on Security is a personal website. Opinions expressed are not necessarily those of Resilient Systems, Inc.