During our departmental "Christmas" dinner, yesterday evening, the subject of an automatic manuscript generator came up. I was kind of intrigued, so I looked it up and found SCIgen, an online tool created by three MIT graduate students.
And of course I couldn't resist.
Studying Operating Systems Using Metamorphic Technology
Albert Einstein, Lotje van der Linden, Sigmund Freud and Sebastiaan Mathot
"Researchers agree that distributed modalities are an interesting new
topic in the field of e-voting technology, and security experts concur.
Such a claim at first glance seems unexpected but fell in line with our
expectations. After years of compelling research into simulated
annealing, we validate the deployment of congestion control, which
embodies the significant principles of e-voting technology. In our
research we introduce a system for 802.11b  (IDE),
arguing that Scheme  and flip-flop gates are entirely
incompatible." [Fulltext PDF]
Kind of cool, right? The manuscript contains graphs and everything. And although the content is totally incoherent (the creators prefer the term "context-free"), there are no screaming grammatical errors.
I was amazed, amused and shocked to learn that a few such generated papers have actually been accepted as conference submissions and by "peer-reviewed" journals. Perhaps it isn't terribly surprising that you can get a paper like that into a shady journal where you pay to publish. But I was shocked that even Elsevier, an established publisher, accepted a random paper for publication in the journal Applied Mathematics and Computation (impact factor 1.124). How on earth can this happen? I think it's hilarious that they even gave a list of suggestions to improve the paper! You can download the article proof here. The fact that the article is actually listed on sciencedirect (retracted, of course) suggests very strongly that this is not a hoax (well, of course it is a hoax, but it's a real one).
On the SCIgen website you can also find the correspondence that followed after the 'authors' admitted to the organizers of the WMSCI 2005 conference that they submitted a bogus paper. Organizers no like. It's worth reading.
You can easily generate your own paper. And who knows, you might even get it published somewhere? (Btw, this form links to the SCIgen website. All credits go to the creators.)