Comments on: Fake text generation the wrong way, and a contest http://www.statisticsblog.com/2013/01/fake-text-generation-the-wrong-way-and-a-contest/ In Monte Carlo We Trust Tue, 21 May 2013 20:58:25 +0000 hourly 1 http://wordpress.org/?v=3.0 By: Matt Asher http://www.statisticsblog.com/2013/01/fake-text-generation-the-wrong-way-and-a-contest/comment-page-1/#comment-16847 Matt Asher Thu, 07 Feb 2013 00:36:18 +0000 http://www.statisticsblog.com/?p=810#comment-16847 I think I may have picked something too obscure. I think I may have picked something too obscure.

]]> By: Paul Matthews http://www.statisticsblog.com/2013/01/fake-text-generation-the-wrong-way-and-a-contest/comment-page-1/#comment-16836 Paul Matthews Mon, 04 Feb 2013 17:13:53 +0000 http://www.statisticsblog.com/?p=810#comment-16836 Nobody else seems to have noticed Matt's "rabbit hole" clue! Nobody else seems to have noticed Matt’s “rabbit hole” clue!

]]> By: Paul Matthews http://www.statisticsblog.com/2013/01/fake-text-generation-the-wrong-way-and-a-contest/comment-page-1/#comment-16833 Paul Matthews Mon, 04 Feb 2013 12:16:05 +0000 http://www.statisticsblog.com/?p=810#comment-16833 Alice in Wonderland. Alice in Wonderland.

]]> By: Alex Foss http://www.statisticsblog.com/2013/01/fake-text-generation-the-wrong-way-and-a-contest/comment-page-1/#comment-16832 Alex Foss Mon, 04 Feb 2013 12:13:12 +0000 http://www.statisticsblog.com/?p=810#comment-16832 Is it _Tacitus and Bracciolini_, by John Wilson Ross? http://www.gutenberg.org/ebooks/9098 Is it _Tacitus and Bracciolini_, by John Wilson Ross?

http://www.gutenberg.org/ebooks/9098

]]> By: Matt Asher http://www.statisticsblog.com/2013/01/fake-text-generation-the-wrong-way-and-a-contest/comment-page-1/#comment-16789 Matt Asher Sat, 26 Jan 2013 20:41:09 +0000 http://www.statisticsblog.com/?p=810#comment-16789 Alright. How about 3 guesses per person per week. Alright. How about 3 guesses per person per week.

]]> By: flodel http://www.statisticsblog.com/2013/01/fake-text-generation-the-wrong-way-and-a-contest/comment-page-1/#comment-16788 flodel Sat, 26 Jan 2013 17:53:46 +0000 http://www.statisticsblog.com/?p=810#comment-16788 Matt, can we change the rules to one guess per person per day? I've looked at the data again and I think I've found interesting things that will help narrow it down... Matt, can we change the rules to one guess per person per day? I’ve looked at the data again and I think I’ve found interesting things that will help narrow it down…

]]> By: Matt Asher http://www.statisticsblog.com/2013/01/fake-text-generation-the-wrong-way-and-a-contest/comment-page-1/#comment-16780 Matt Asher Fri, 25 Jan 2013 20:19:27 +0000 http://www.statisticsblog.com/?p=810#comment-16780 Nope. Tks for sharing your procedure. Nope. Tks for sharing your procedure.

]]> By: Jonas http://www.statisticsblog.com/2013/01/fake-text-generation-the-wrong-way-and-a-contest/comment-page-1/#comment-16772 Jonas Thu, 24 Jan 2013 14:19:14 +0000 http://www.statisticsblog.com/?p=810#comment-16772 My guess is The Life and Adventures of Robinson Crusoe by Daniel Defoe. I fitted a first order markov model for the generated text, scored the topp100 at project gutenberg against the fitted transition matrix and selected the one with highest log probability. My guess is The Life and Adventures of Robinson Crusoe by Daniel Defoe. I fitted a first order markov model for the generated text, scored the topp100 at project gutenberg against the fitted transition matrix and selected the one with highest log probability.

]]> By: Matt Asher http://www.statisticsblog.com/2013/01/fake-text-generation-the-wrong-way-and-a-contest/comment-page-1/#comment-16765 Matt Asher Thu, 24 Jan 2013 02:04:47 +0000 http://www.statisticsblog.com/?p=810#comment-16765 Nice guess! That would have been a good one to use, but it wasn't the one I did use. I'm not sure if this is an easy contest to solve or if it will send people down the (beautiful?) rabbit hole of support vector machines and similar. Good luck to all either way. Nice guess! That would have been a good one to use, but it wasn’t the one I did use.

I’m not sure if this is an easy contest to solve or if it will send people down the (beautiful?) rabbit hole of support vector machines and similar. Good luck to all either way.

]]> By: flodel http://www.statisticsblog.com/2013/01/fake-text-generation-the-wrong-way-and-a-contest/comment-page-1/#comment-16764 flodel Thu, 24 Jan 2013 01:27:29 +0000 http://www.statisticsblog.com/?p=810#comment-16764 Webster's Unabridged Dictionary. Webster’s Unabridged Dictionary.

]]>