Comments on: Seeing how “art” and “pharmaceuticals” are linguistically similar in web text http://brenocon.com/blog/2009/09/seeing-how-art-and-pharmaceuticals-are-linguistically-similar-in-web-text/ cognition, language, social systems; statistics, visualization, computation Tue, 25 Nov 2025 13:11:20 +0000 hourly 1 By: Bryan Routledge http://brenocon.com/blog/2009/09/seeing-how-art-and-pharmaceuticals-are-linguistically-similar-in-web-text/#comment-24269 Bryan Routledge Wed, 31 Mar 2010 02:02:58 +0000 http://anyall.org/blog/?p=745#comment-24269 Very cool. (nothing deeper to add; just very cool)

]]> By: brendano http://brenocon.com/blog/2009/09/seeing-how-art-and-pharmaceuticals-are-linguistically-similar-in-web-text/#comment-15763 brendano Sat, 07 Nov 2009 17:19:42 +0000 http://anyall.org/blog/?p=745#comment-15763 thanks, fixed

]]> By: Andy Carlson http://brenocon.com/blog/2009/09/seeing-how-art-and-pharmaceuticals-are-linguistically-similar-in-web-text/#comment-15749 Andy Carlson Sat, 07 Nov 2009 13:22:36 +0000 http://anyall.org/blog/?p=745#comment-15749 Nice post, Brendan. I wonder what the lists would look like if you replaced words with Wordnet synsets to make the “occurs with both art and pharmaceuticals” criterion less dependent on specific words chosen.

A minor correction: that should be “filtered to contexts that appear five hundred times in the data,” not “five hundred times for the noun phrase.” Co-occurrence counts can be as small as one.

]]>