Wouldn't a mechanical Google search, using a DTM (Document Term Matrix)
give at least a human-in-the-loop system a good chance of finding the
frauds?

One obvious fraud would be to submit the paper in a different language than
the original, right?  But even so, DTM approaches would likely work on
Google due to their multi-language search capability.

   -- Owen


On Wed, May 29, 2013 at 5:30 PM, Roger Critchlow <[email protected]> wrote:

> I don't know about re-submission detection.  A current brouhaha noted in
> today's Nature,
> http://www.nature.com/news/fallout-from-hailed-cloning-paper-1.13078, concerns
> a paper about cloning skin cells into stem cells which was granted
> expedited review and on-line publication by Cell, one of the premier
> journals.   So, "an anonymous online commenter noted three pairs of
> duplicated images with conflicting labels in the paper", but none of the
> paid editorial staff or reviewers had happened to notice these problems in
> the 3 days it took them to review it or the 12 subsequent days it took to
> prepare it for online publication.
>
> Rockefeller University Press has a system to automatically check that the
> submitted graphics are not duplicates of each other, but most journals
> still do manual spot checking (Nature checks 2 articles per issue on
> average) if they do anything at all.  Checking for duplicate submissions or
> more subtle plagiarism is beyond their abilities, they're too busy
> "curating" the scientific literature.
>
> Back to the original post, that the reviewers basically slammed papers
> that they had already published is pretty sickening.  The system has
> nothing to prevent such crony reviewing claques even now.
>
> -- rec --
>
>
> On Wed, May 29, 2013 at 4:53 PM, Russell Standish 
> <[email protected]>wrote:
>
>> On Wed, May 29, 2013 at 02:14:26PM -0600, Roger Critchlow wrote:
>> > Russ Abbott reposted this on g+, but it's too meritorious not to be
>> > archived here:
>> >
>> >
>> >
>> http://journals.cambridge.org/action/displayAbstract?fromPage=online&aid=6577844
>> >
>> > Take published articles by highly respected authors, replace the authors
>> > and institutions with fakes, resubmit to the same journals that
>> originally
>> > published the articles, and watch what happens.
>> >
>>
>> What astounded me was the very low detection of resubmission
>> (8%). This was in spite of the articles having been published within
>> the previous two years in the _same_ journals. Obviously these must be
>> large journals with multiple editors who clearly aren't across what
>> their colleagues are doing.
>>
>> The other concerning thing is that the rejection rate of the papers
>> that pass that filter is so high (89%), particularly that it is higher
>> than the rejection rate for new articles submitted to. Obviously, I would
>> expect
>> the rejection rate to be greater than 0, but it should be less than
>> the overall journal rejection rate, as the paper concerned have
>> already run the gauntlett of peer review. I guess the conclusion is
>> that referees were being influenced by who wrote the paper, not the
>> contents.
>>
>> It'd be interesting to redo the experiment today - although I would
>> hope that journals now have a better detection of article resubmission in
>> place.
>>
>> Cheers
>>
>> --
>>
>>
>> ----------------------------------------------------------------------------
>> Prof Russell Standish                  Phone 0425 253119 (mobile)
>> Principal, High Performance Coders
>> Visiting Professor of Mathematics      [email protected]
>> University of New South Wales          http://www.hpcoders.com.au
>>
>> ----------------------------------------------------------------------------
>>
>> ============================================================
>> FRIAM Applied Complexity Group listserv
>> Meets Fridays 9a-11:30 at cafe at St. John's College
>> to unsubscribe http://redfish.com/mailman/listinfo/friam_redfish.com
>>
>
>
> ============================================================
> FRIAM Applied Complexity Group listserv
> Meets Fridays 9a-11:30 at cafe at St. John's College
> to unsubscribe http://redfish.com/mailman/listinfo/friam_redfish.com
>
============================================================
FRIAM Applied Complexity Group listserv
Meets Fridays 9a-11:30 at cafe at St. John's College
to unsubscribe http://redfish.com/mailman/listinfo/friam_redfish.com

Reply via email to