Comments on: Spam stats http://bit-player.org/2008/spam-stats An amateur's outlook on computation and mathematics. Mon, 01 Dec 2008 20:50:52 +0000 http://wordpress.org/?v=2.6.3 By: اموزش http://bit-player.org/2008/spam-stats#comment-1744 اموزش Fri, 25 Jul 2008 15:41:34 +0000 http://bit-player.org/?p=148#comment-1744 VERY GOOD VERY GOOD

]]> By: بدن ساز http://bit-player.org/2008/spam-stats#comment-1704 بدن ساز Sat, 21 Jun 2008 09:46:33 +0000 http://bit-player.org/?p=148#comment-1704 Thanks Thanks

]]> By: Jim Ward http://bit-player.org/2008/spam-stats#comment-1699 Jim Ward Fri, 06 Jun 2008 13:00:46 +0000 http://bit-player.org/?p=148#comment-1699 Speaking of science and sex, Mary Roach has a new book out, "Bonk". Her previous books, "Stiff" and "Spook" were pretty good. I only have one recommendation for a group theory book ... Speaking of science and sex, Mary Roach has a new book out, “Bonk”. Her previous books, “Stiff” and “Spook” were pretty good. I only have one recommendation for a group theory book …

]]> By: Jim Ward http://bit-player.org/2008/spam-stats#comment-1698 Jim Ward Fri, 06 Jun 2008 12:44:57 +0000 http://bit-player.org/?p=148#comment-1698 Re: The "mistaken for birdshit" link, I was listening to Feynman's "What Do You Care What Other People Think?" yesterday and I was interested to learn that in his day physics was divided between experimentalists and theorists. The experimentalists would discover some odd fact, and then go to the theorists for an explanation. It looks like June 1963 was the exact point when the experimentalists and theorists caught up with each other. Today the theorists have lapped the experimentalists, which explains the slow progress of physics today. Too much theory, not enough data. I also learned of tang and clevis joints, in case you want to write "Physics in the Bedroom". De Sade has beaten you to "Philosophy". Re: The “mistaken for birdshit” link, I was listening to Feynman’s “What Do You Care What Other People Think?” yesterday and I was interested to learn that in his day physics was divided between experimentalists and theorists. The experimentalists would discover some odd fact, and then go to the theorists for an explanation. It looks like June 1963 was the exact point when the experimentalists and theorists caught up with each other. Today the theorists have lapped the experimentalists, which explains the slow progress of physics today. Too much theory, not enough data.

I also learned of tang and clevis joints, in case you want to write “Physics in the Bedroom”. De Sade has beaten you to “Philosophy”.

]]> By: brian http://bit-player.org/2008/spam-stats#comment-1696 brian Thu, 05 Jun 2008 19:42:00 +0000 http://bit-player.org/?p=148#comment-1696 I was hoping you would ask. It's such a fascinating story! I counted those emails by hand, scanning 2,715 subject lines in chronological order. The first non-English message I came to was in Russian, so I listed that category first. Next came a message in an Asian language. Do you begin to see the pattern? The first German email happened to arrive <em>after</em> the first Italian one but <em>before</em> the first Spanish one. (The "unknown" category is something I split off from the Asian group after the fact; I'm pretty sure those messages are in some form of Japanese.) By the way, I never want to do this kind of counting again. I had tried to automate the language labelling by extracting strings such as "charset=iso-2022-jp" (which designates a Japanese character encoding). The results were totally bogus. Most of the messages have no such encoding designators, and many of them have misleading markers, or malformed ones. Next time -- if there is a next time -- I'm going to writing a little <em>n</em>-gram language recognizer. Or else figure out how to abuse the Google language tools and have them do the job for me. I was hoping you would ask. It’s such a fascinating story!

I counted those emails by hand, scanning 2,715 subject lines in chronological order. The first non-English message I came to was in Russian, so I listed that category first. Next came a message in an Asian language. Do you begin to see the pattern? The first German email happened to arrive after the first Italian one but before the first Spanish one. (The “unknown” category is something I split off from the Asian group after the fact; I’m pretty sure those messages are in some form of Japanese.)

By the way, I never want to do this kind of counting again. I had tried to automate the language labelling by extracting strings such as “charset=iso-2022-jp” (which designates a Japanese character encoding). The results were totally bogus. Most of the messages have no such encoding designators, and many of them have misleading markers, or malformed ones.

Next time — if there is a next time — I’m going to writing a little n-gram language recognizer. Or else figure out how to abuse the Google language tools and have them do the job for me.

]]> By: Barry Cipra http://bit-player.org/2008/spam-stats#comment-1695 Barry Cipra Thu, 05 Jun 2008 16:32:28 +0000 http://bit-player.org/?p=148#comment-1695 Just curious: How did you choose the ordering of the wedges in your pie charts? In particular, why place German between two romance languages? Just curious: How did you choose the ordering of the wedges in your pie charts? In particular, why place German between two romance languages?

]]>