Bayesian Storystarter

Most know that spammers are now attempting to defeat Bayesian auto-classifiers by including either mass amounts of dictionary words, or otherwise randomly generated text, in their emails. I received the following blurb today, which sounds like the start of a pretty good story. know the real text this was based off?

And how do you treat these marvelous gifts? Why, you carry them to a cannibal island, where even your crude civilization has not yet penetrated! I wanted to astonish the natives, said Rob, grinning

The Demon uttered an exclamation of anger, and stamped his foot so fiercely that thousands of electric sparks filled the air, to disappear quickly with a hissing, crinkling soundYou might have astonished those ignorant natives as easily by showing them an ordinary electric light, he cried, mockingly The power of your gifts would have startled the most advanced electricians of the worldWhy did you waste them upon barbarians? Really, faltered Rob, who was frightened and awed by the Demon's vehement anger, I never intended to visit a cannibal island

I meant to go to Cuba Cuba! Is that a center of advanced scientific thought? Why did you not take your marvels to New York or Chicago; or, if you wished to cross the ocean, to Paris or Vienna? I never thought of those places, acknowledged Rob, meeklyThen you were foolish, as I said, declared the Demon, in a calmer tone

Oogling for phrases returns The World Key by L. Frank Baum (author of THE WIZARD OF OZ): "an Electrical fairy tale founded upon the mysteries of electricity and the optimism of its devotees. It was written for boys, but others may read it."

"To be sure. A demon is also a genius; and a genius is a demon," said the Being. "What matters a name? I am here to do your bidding."

Regardless, it appears this is nothing new, as the BBC reported in December:

But as Clive Thompson points out, automatically generating text that reads like it was written by a human hand is difficult. This is perhaps why some spammers are turning to out-of-copyright novels for their text. It is an ideal source of real writing.