Most of this data is completely useless, because there's no way to tell where these users live, at least without a serious amount of stalking and sleuthing.
Unless, of course, they've been geotagging their tweets from a mobile device. In which case it would then be trivial to look at their most frequent location, determine it's a residential one, wait until they're out, and head on over.
Which is why I will never enable Twitter geotagging.
I searched for my city, and found the most recent listing on the site. Clicked through to that Twitter user's profile, then to their website which was about yoga classes. On her yoga website, she lists her home address. It's in the WHOIS data of her domain too.
Chances are it's not that rare to be able to get an address for someone that's connected enough to be tweeting their location all the time.
Does the web site also list class schedules? Because that is also a source of “when she is not at home”, as someone else already said, most people who have a job are out during the day, so finding “empty homes” is not exactly rocket science.
Also, in addition to “empty home” you also want a place where no-one hears you if you break a window or similar, a place where it is easy to leave with the stolen goods without being noticed, and a place that has stuff worth stealing.
Given the variables, were I a professional thief, I would scout for locations first, then find out when the residents were not at home second, which is probably fairly simple even without the use of Twitter.
Take it even further. Scout the location, observe the 'regular' schedule (of leaving for work/school) and then use Twitter to make certain they are not close to home.
You can't take too much time observing, since you will probably want a better hourly wage than working at Burger King when you are going to rob someone.
Given that google acts as a phonebook, simply googling someone by real name has a good chance of digging up their address. I picked a friend in the US at random, simply knowing which state he lived in, and I found his phone number, address, and ZIP code. Also a handy link to google maps.
People are no where near as isolated as they'd like to believe.
Agreed. What would make this app REALLY insteresting is if it tried to automatically analyze all of a users geotagging, guess a home/work location, and then alert when the user is over an hour or two away.
Twenty minutes to make sure no one is in the house followed by a quick peek through the mail box (or windows) will give you a pretty solid idea if they do, though.
Unless, of course, they've been geotagging their tweets from a mobile device. In which case it would then be trivial to look at their most frequent location, determine it's a residential one, wait until they're out, and head on over.
Which is why I will never enable Twitter geotagging.