You might have heard a rather surprising news on a discovery made by searching the publications database. This is the headtitle as it appeared in Scientific American on Jan 30, 2008: “Database search turns up research papers suspiciously similar to prior publications, prompting investigations“.
This finding is probably not new and it was perhaps somehow known in the field, but not the extension, the size of it!, which I personally find the most astonishing part of the history. They seem to have found literally thousands of articles which appear to be too similar to other previous publications. And they only have scanned the abstracts!
Researchers Mounir Errami and Harold “Skip” Garner of the University of Texas Southwestern Medical Center at Dallas used a text-matching algorithm (their own software tool, called eTBLAST) to compare millions of Medline abstracts uncovering those identified as being closely related.