"Jonathan M Davis" <jmdavisp...@gmx.com> wrote in message news:mailman.2200.1299299436.4748.digitalmar...@puremagic.com... > On Friday 04 March 2011 19:48:10 uri wrote: >> This should likely go to d.learn or someplace else. As a simple exercise >> I >> tried to build a tool that fetches a list of random news articles written >> by a single person and determine his or hers time zone based on the >> frequencies. >> >> I first thought the period of sleep would determine the daily rhythm, but >> that doesn't seem to be the case. In http://piczasso.com/i/amj97.png I >> used bearophile's data points and the sleep period gave GMT-4 .. GMT-3, >> dinner time GMT+0 .. GMT+2, GMT+1 having the highest probability, which >> is >> the correct answer. I wish U sleep all well and don't work too hard 24/7, >> the AI is watching U. > > You're assuming that we have normal sleep patterns. ;) >
A program like that would probably conclude that I jet-set between Hawaii and Japan rather than living in Cleveland. It's a *very* interesting idea, though. Might work great for other disciplines. Although programmers tend to be the exceptions to many rules. :) Another interesting twist might be to run a spellchecker on the posts (but first attempt to recognize code and psuedo-code and omit it from the spellchecking), with the idea that more errors from the same person might imply more fatigue (or maybe just more busy during the workday).