News for hackers. Stuff that matters.
Re:Using Gowalla checkins to predict alcohol-related travel by chronomex on Tuesday March 22, 2011 @03:06AM (#2353440)
This was my final project for the quarter. I was surprised at how many Gowalla profiles were matched exactly by voter registrations.
Re:Using Gowalla checkins to predict alcohol-related travel by jrockway on Tuesday March 22, 2011 @03:42AM (#2353511)
Some very interesting results, especially that only 20 out of 1000+ people sampled made their names on Facebook private.

Something that I thought was strange is that I usually go out drinking near work rather than near home, and then ride my bike or take the bus home afterwards. I also routinely walk more than a mile. The article seems to consider this uncommon, but it's common amongst my group of city-dwelling friends.

It would also be interesting to use this data to see how public transportation affects drinking. When I go drinking at a bar that's near the Green Line, which closes around 1:20am, I make sure to finish up and leave around 1 (unless the extra hour is really going to make a difference to me). When I go drinking near a bar that's near the Red Line, which is 24 hours, I rarely leave before closing time. Am I normal? Mining the data could tell me :)

Re:Using Gowalla checkins to predict alcohol-related travel by chronomex on Tuesday March 22, 2011 @04:09AM (#2353551)
I thought I had my Facebook set to maximal privacy, but it turned out that I didn't. I'm not too surprised at this, to be honest.

You think you can do something else cool with the data? Have at it! :) Let me know if other data such as more detailed user profile info will help.

http://xrtc.net/f/projects/gowallahol/gowalla_checkins.csv.g...

Re:Using Gowalla checkins to predict alcohol-related travel by expertio on Tuesday March 22, 2011 @03:40AM (#2353509)
I have to say this is so cool! Police department could use this data to do roadside random breath test. And there will be less killer on the road.
Re:Using Gowalla checkins to predict alcohol-related travel by chronomex on Tuesday March 22, 2011 @04:11AM (#2353554)
Thanks! Our original intention was to do exactly that, though we couldn't figure out how to do route planning with ArcGIS for more than one origin-destination pair at a time.
Re:Using Gowalla checkins to predict alcohol-related travel by jrockway on Tuesday March 22, 2011 @03:42AM (#2353513)
Probably not. The number of people driving drunk is much lower than the number of people driving distracted or driving stupid.
Re:Using Gowalla checkins to predict alcohol-related travel by evolvingstuff on Tuesday March 22, 2011 @05:38AM (#2353664)
There may indeed be more distracted/stupid people, but the inebriated ones cause about twice the number of deaths overall in the US:

http://www.cdc.gov/MotorVehicleSafety/distracted_driving/ind...

http://www.cdc.gov/MotorVehicleSafety/Impaired_Driving/impai...