posted by
janrinok
on Monday March 03 2014, @09:30PM
from the twas-brillig-and-the-slithy-toves dept.
from the twas-brillig-and-the-slithy-toves dept.
AnonTechie writes:
"The publishers Springer and IEEE are removing more than 120 papers from their subscription services after a French researcher discovered that the works were computer-generated nonsense.
Over the past two years, computer scientist Cyril Labbe of Joseph Fourier University in Grenoble, France, has catalogued computer-generated papers that made it into more than 30 published conference proceedings between 2008 and 2013. Sixteen appeared in publications by Springer, which is headquartered in Heidelberg, Germany, and more than 100 were published by the Institute of Electrical and Electronic Engineers (IEEE), based in New York. Both publishers, which were privately informed by Labbe, say that they are now removing the papers.
This discussion has been archived.
No new comments can be posted.
Publishers Withdraw More than 120 Gibberish Papers
|
Log In/Create an Account
| Top
| 40 comments
| Search Discussion
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
(Score: 4, Funny) by omoc on Monday March 03 2014, @09:39PM
So either the people in peer review don't do their jobs and just pass things along and or they were impressed by the number of buzzwords so they just had to wave these through?
(Score: 5, Insightful) by BsAtHome on Monday March 03 2014, @09:44PM
Probably the former. Doing peer review is actually a lot of work and that is long down on the priority list and most reviewers are not paid for the peer-review. The selection of reviewers is for some journals also slightly arbitrary.
(Score: 5, Informative) by bd on Monday March 03 2014, @10:13PM
For many conferences, proceedings are actually not peer-reviewed at all. They just check for the format. They think it would be embarassing enough if you put your name next to gibberish that will be associated with you henceforth.
On conferences, you usually report on what you are doing right now, what the problems are and what new stuff you are working on, so that others in your field know what you are doing. You try to do that without revealing what you have not yet published. The proceeding is supposed to summarize your talk, but often written before you even know exactly what you will say. It sometimes is seen as a hassle by many and some people think no one will ever read it.
When you make an actual scientific discovery, you publish a proper paper, that is why you usually cite peer-reviewed papers rather then conference proceedings.
I think I read an abstract a while back that was about the applications of dilithium crystals in warp engines.
(Score: 4, Insightful) by weeds on Monday March 03 2014, @10:28PM
That may be true, but from the article:
So those "reviewers" have some explaining to do.
Get money out of politics! [mayday.us]
(Score: 4, Interesting) by bd on Monday March 03 2014, @11:25PM
What do you expect the publisher to say ;-). Fact of the matter is, it is already well known that "peer-review" for proceedings is - let's say - not very rigorous. Actually, IEEE proceedings are sometimes one page of abstract and a page filled with figures not related to the abstract, that seem like the author had flying around on his desk. I don't know whether humans read them, I have until now never got interesting information out of one. Computer generated or not.
But you are right, someone was probably given the designation "peer-reviewer" for each of the sessions these talks were in, and I would not want to be him right now.
(Score: 2) by weeds on Tuesday March 04 2014, @01:29AM
Yeah, they seem to be backing themselves into a corner don't they?
Get money out of politics! [mayday.us]
(Score: 2) by TheRaven on Tuesday March 04 2014, @10:24AM
sudo mod me up
(Score: 1) by goose on Tuesday March 04 2014, @07:58AM
This is not the first time that a scientific publisher got trolled (nor would it be the last). The sad reality is that getting published has little to do with actual new research. For example, I wrote this - safe link [lelanthran.com] - but have no place to submit for publication. All the places that have an audience who might be interested have strict guidelines about the submitter.
(Score: 2) by VLM on Monday March 03 2014, @11:02PM
"They just check for the format."
I've seen enough conference proceedings to know they don't even bother with that. Typographical / graphics arts epic fails.
(Score: 1) by lx on Tuesday March 04 2014, @07:23AM
Dammit Jim, I'm a doctor not a typographer!
(Score: 1) by axsdenied on Tuesday March 04 2014, @03:23PM
I have reviewed proceedings at a conference in the past (Physics). We were given the papers on the 2nd day and asked to finish reviewing before the end of the conference which in total lasted 3 days.
The best I could do in the limited time was to read the paper and comment on the obvious things. There was no time to follow references and read more about the topic. But I doubt most of people even bothered to do that much...
(Score: 2) by AnonTechie on Tuesday March 04 2014, @07:23AM
Haven't they outsourced peer review to one of the third world countries ??
Albert Einstein - "Only two things are infinite, the universe and human stupidity, and I'm not sure about the former."
(Score: 1) by Rivenaleem on Tuesday March 04 2014, @09:52AM
To be peer reviewed, they would have to have been reviewed by computer too. And to a computer the papers may have seemed perfectly fine, perhaps even intelligible. Computers might be becoming aware, but incredibly dumb. It's not a surprise considering their primary function seems to be putting captions on cat pictures.
(Score: 4, Funny) by theluggage on Monday March 03 2014, @09:50PM
Wow, exciting moment. This page is required to post a comment. Unfortunately, that's fine, but not to me the gate of the designers need more. No, it's elit free, a quiver organization.
It's a great advantage of the course , or chocolate consumption and enhanced. This page is required to post a comment. Lake visitors live performances or organizational official website of the author of the pain in my throat. It's a great advantage of the course , or chocolate consumption and enhanced. I really had a job . Apple patients than playing the game of football betting. Welcome to Oklahoma but no drink driving . Chinese cultural beliefs.
It's a soft , there is no advantage of the grief, but it was the ferry around the world, a lot of the development on the fringe of hatred nor the meals. How to Choose protein, development and of course advantages, to create a macro to improve the budget , in order to warm-up the mass of the option might be a bit tight . Unfortunately, that's fine, but not to me the gate of the designers need more . I really had a job . Apple patients than playing the game of football betting . How to Choose protein, development and of course advantages , to create a macro to improve the budget , in order to warm-up the mass of the option might be a bit tight.
(Score: 3) by gishzida on Monday March 03 2014, @10:05PM
I had nothing to do with the parent post but everything to do with this: http://markovswisdom.blogspot.com/ [blogspot.com]
That said, for all the talk of "the [coming real soon now, the new improved] A.I. singularity" computers are still stupid. How will we know they are smart? When they learn where to buy their term papers.
(Score: 3, Funny) by dotdotdot on Monday March 03 2014, @10:13PM
You can't fool me. There is no way a computer could come up with this ...
(Score: 2, Informative) by gishzida on Monday March 03 2014, @10:59PM
The input text steams that were used for that particular output run were:
"the book of reminders" [an as yet unpublished book of quotations gleaned from my years of endeavors at the advanced school of hard knocks] and a stream of Lewis Carroll's writings...
But among the quotes in the book of reminders is:
"Magic is the ability to suspend our disbelief: Play with a cat. Watch a child. Look in the Mirror. Magic is REAL."
And a quote about spices: "Always set aside a portion of you income for spices. Spices can make even the dullest life liveable."
Hench the Markov Wisdom of "Always set aside a cat" is likely to be true.
Thanks for reading.
(Score: 1) by theluggage on Tuesday March 04 2014, @12:10AM
Wow - that's sophisticated.
I just plugged the output of a "lorem ipsum" utility [littleipsum.com] into Google Translate. And I got modded +5! so just think what could be achieved with a really sophisticated posting generator.
Come to think of it... hello...? is there actually anybody out there?
(Score: 1) by gishzida on Tuesday March 04 2014, @12:54AM
Who do you think is doing the moderation? It's what you get for taking one of those funny colored capsules they talk about in the movies. You never know where you'll wake up.
(Score: 1) by Taibhsear on Tuesday March 04 2014, @06:22PM
Usually, the local drunk tank.
(Score: 4, Interesting) by Ethanol-fueled on Monday March 03 2014, @10:15PM
Hahah, nice.
Your post reads a lot like Atlanta Nights, [wikipedia.org] a collaborative effort written by a bunch of pissed-off Sci-Fi writers trolling a publishing house because it hosted articles basically saying that Sci-fi is crap and Sci-fi writers are a bunch of illiterate neckbeards. From the article:
" The distinctive flaws of Atlanta Nights include nonidentical chapters written by two different authors from the same segment of outline (13 and 15), a missing chapter (21), two chapters that are word-for-word identical to each other (4 and 17), two different chapters with the same chapter number (12 and 12), and a chapter "written" by a computer program that generated random text based on patterns found in the previous chapters (34).
After PublishAmerica accepted the manuscript for publication, the hoax was revealed by the authors. I have this book, and man, is it hilariously terrible -- it's just like a Slashdot discussion turned into a novel, with misused apostrophes and missing closing-quotes, loaded with the kind of dumb sayings you'd expect to hear from Philip J. Fry. From the first page in the opening chapter:
Pain. Whispering voices. Pain. Pain. Pain. Pain.
Need pee--new pain--what are they sticking in me? . . .
Sleep.
Pain.
Whispering voices.
"As you know, Nurse Eastman, the government spooks controlling this hospital
will not permit me to give this patient the care I think he needs."
"Yes, doctor." The voice was breathy, sweet, so sweet and sexy..."
Etc.
(Score: 2, Informative) by etherscythe on Tuesday March 04 2014, @12:52AM
Slightly more to the point (I sat in on a presentation by several Science Fiction & Fantasy Writers of America [sfwa.org] members on this topic), the PublishAmerica company was basically recruiting authors with promises of advertising, editing, and other support which they completely failed to provide after they received their money. When authors would complain, they would respond in derogatory fashion, at which point that book concept was born. It showed up the highly-touted "publication standards" of the company in pretty epic fashion.
"Fake News: anything reported outside of my own personally chosen echo chamber"
(Score: 2, Interesting) by Anonymous Coward on Tuesday March 04 2014, @03:09AM
A solution would be for each peer reviewer to sign their name to the publication that they peer reviewed it. If we see that a peer reviewer signs off on way more articles than they can possibly read and reasonably understand, analyze, scrutinize, criticize, and investigate then we know something is suspicious. Peer reviewers should only be permitted to sign off on a limited, reasonable, number of publications a year in opposed to simply rubber stamping 20 a week (made up number) that they couldn't have possibly even gone through (yet alone thoroughly) if you consider all their other routine daily activities that they must also spend time doing (eating, sleeping, do they have another job and how much time is spent there, how much time do they allot to peer reviewing and is it on site monitored peer reviewing, where their time can be audited, or is it at home reviewing, etc..).
(Score: 2, Interesting) by Anonymous Coward on Tuesday March 04 2014, @03:18AM
(same author, sorry for so many submissions). Another possible solution could be to somehow require the person peer reviewing the document to answer basic questions and to write those questions down.
Summarize the article. State its conclusions. Does the article prescribe anything? If so, what? How was this conclusion arrived at? What was the experiment done and what is the procedure? What is the uncertainty? What measurements were made and how were they made?
I'm sure others can think of more difficult questions.
If they can't even be bothered to answer some basic questions then they don't deserve to be selected to peer review anything.
(Score: 0) by Anonymous Coward on Tuesday March 04 2014, @03:21AM
(errrr ... write those answers down. They don't get to choose the questions).
(Score: 0) by Anonymous Coward on Tuesday March 04 2014, @02:59AM
I nominate you for a noble prize!!!
(hey, Obama got one ...)
(Score: 2) by Toaster42 on Monday March 03 2014, @09:58PM
It has all the stylings of a hyperlink, but in my browser (Pale Moon 24.3.2 (x64)) clicking it doesn't do much of anything.
Is it just me?
All higher forms of thinking come from neural connections built by solving the kinds of problems encountered in math.-Md
(Score: 2, Insightful) by fishybell on Monday March 03 2014, @10:02PM
It's not just you; there is no href in the tag.
From the page source: <a>removing more than 120 papers from their subscription services</a>
(Score: 5, Funny) by istartedi on Monday March 03 2014, @10:00PM
Will there be any effort to revoke the computer's tenure?
Appended to the end of comments you post. Max: 120 chars.
(Score: 1, Redundant) by dtremenak on Monday March 03 2014, @10:01PM
It's always good to keep up Slashdot's stellar editing conventions. The article link doesn't actually go anywhere (it's an empty <a> tag).
(Score: 4, Informative) by song-of-the-pogo on Monday March 03 2014, @10:04PM
I'm guessing, based on a highly similar post at "that other site", the link was meant to be:
http://www.nature.com/news/publishers-withdraw-mor e-than-120-gibberish-papers-1.14763 [nature.com]
"We have met the enemy and he is us."
(Score: 2, Informative) by barrahome on Monday March 03 2014, @10:12PM
Here you have sir! http://www.nature.com/news/publishers-withdraw-mor e-than-120-gibberish-papers-1.14763 [nature.com]
(Score: 1) by skullz on Monday March 03 2014, @11:19PM
Obviously the S/N editors also moonlight as conference paper reviewers. Now we too can experience top academic quality news stories.
(Score: 4, Interesting) by song-of-the-pogo on Monday March 03 2014, @10:52PM
Btw, here's the link to SCIGen: http://pdos.csail.mit.edu/scigen/ [mit.edu]
The paper it generated for me is titled "The Effect of Stable Modalities on Software Engineering" and the abstract reads as follows:
"We have met the enemy and he is us."
(Score: 3, Interesting) by Khyber on Monday March 03 2014, @10:54PM
they publish so many bullshit psychological papers that they might as well be ruled as totally unacceptable for research in any shape or form.
They're so full of shit they can't explain my own social issues, and they've tried.
So, yea, don't trust Springer/IEEE. They're likely NSA-run.
Destroying Semiconductors With Style Since 2008, and scaring you ill-educated fools since 2013.
(Score: 2, Funny) by gringer on Monday March 03 2014, @11:26PM
Many theorists would agree that, had it not been for Boolean logic, the evaluation of gibberish papers through checksums might never have occurred. Given the current status of autonomous archetypes, scholars dubiously desire the study of the lookaside buffer as an alternative to checksum analysis. Such a hypothesis is generally an appropriate ambition but is derived from known wrong results, namely slashdot posts.
Our algorithm relies on the practical architecture outlined in the recent infamous work by Martinez et al. in the field of steganography; this seems to hold in most cases. We postulate that modular technology can manage atomic comparisons of gibberish without needing to learn stochastic algorithms. Each component of our algorithm, UlmicOrf, runs in O(n) time, independent of all other components. The architecture for our heuristic consists of four independent components: encrypted algorithms, the Internet, collaborative social networking websites, and RFCs. Thusly, despite substantial work in this area, our method is apparently the application of choice among experts
Ask me about Sequencing DNA in front of Linus Torvalds [youtube.com]
(Score: 1) by martink on Monday March 03 2014, @11:56PM
(Score: 0) by Anonymous Coward on Tuesday March 04 2014, @01:04AM
A: have a computer generate gibberish
B: Submit it
C: Nobody will read it
D: It gets approved
F: It gets published
G: ???
H: PROFIT!!!!!
(Score: 1) by _NSAKEY on Tuesday March 04 2014, @02:52AM
Will some enterprising comedian start a satirical journal with these gibberish papers?
(Score: 1) by photong on Tuesday March 04 2014, @07:04AM
As someone who proofreads engineering papers for a living, this is not surprising. Reviewers range from being really thorough and helpful to just plain silly. One paper can have one reviewer state "I strongly recommend this paper for publication as the findings are novel and useful" and another state "This paper cannot be published in its current state" followed by a long list of problems. Reviewing papers is just not sexy enough for most people to care enough.