Print

Print


New development! Here's something I think the same guy wrote
http://www.lycaeum.org/mv/M2/woowoo.html

On Tue, Oct 13, 2009 at 11:33 PM, Alex Fink <[log in to unmask]> wrote:

> On Tue, 13 Oct 2009 10:38:00 -0700, Gary Shannon <[log in to unmask]>
> wrote:
>
> >Collecting stats on several different pages it appears that the
> >accents are added randomly after the words are generated. Word
> >distribution statistics look more uniform when all the accents are
> >removed.
>
> No, I think it's generated with the accents in place.  The sequence <o>
> appears, for instance, but I've never seen a plain <>.
>
> >It's certainly not a language. I looked at a bunch of different
> >natural language texts and collected one simple statistic: number of
> >unique words divided by total number of words in the document. The
> >natural languages I looked at varied between a ratio of 0.071 and
> >0.110. The gibberish samples were all off the charts at around 0.692
> >to 0.711 putting it well outside the range of natural languages.
>
> That's similar to what Jim found, then, not enough repeated words.  My
> guess
> would be that each word is being picked independently and context-freely
> from the same distribution.
>
> Alex
>