{"version":"1.0","provider_name":"Rainbow Dash Network","provider_url":"http:\/\/rainbowdash.net\/","type":"link","title":"Bit Shift (bitshift)'s status on Thursday, 07-Jun-12 04:05:45 UTC","author_name":"Bit Shift (bitshift)","author_url":"http:\/\/rainbowdash.net\/bitshift","url":"http:\/\/rainbowdash.net\/notice\/1503684","html":"@<span class=\"vcard\"><a href=\"http:\/\/rainbowdash.net\/user\/9619\" class=\"url\" title=\"Mike M\"><span class=\"fn nickname\">techdisk42<\/span><\/a><\/span> Well, @<span class=\"vcard\"><a href=\"http:\/\/rainbowdash.net\/user\/7201\" class=\"url\" title=\"anypony\"><span class=\"fn nickname\">anypony<\/span><\/a><\/span> itself gets it from <a href=\"http:\/\/twitter.com\/horse_ebooks\" title=\"http:\/\/twitter.com\/horse_ebooks\" rel=\"nofollow external\">twitter.com\/horse_ebooks<\/a>. As for how that does it; while I don't know the specific method, I'm pretty sure it's based on statistical analysis of word frequency in a sample text, and then generating new text that follows the same patterns. Markov chains are a fairly simple-to-understand example, if you want to look into this more, though I don't think it's the exact variety used by that bot."}