Hey. This is a gradual attempt to turn Georgia's CSV sex offender
registry into an automagically updated Google Fusion Tables map. It's a work
in progress, to be sure. A to-do list appears in the main file, scrape.py,
and may even be sort of kind of up-to-date.

scrape.py also includes descriptions of what the fields are and where they
come from. It also includes sample code for Fusion Tables.

Use easy_install to fetch geopy. Everything else should come stock with
Python 2.7.

Scream if you need me. Collaborators welcome, though I have no earthly
idea who else would have an interest in this.	

Yeah, this data is almost certainly awful. Problems with using a sex offender registry for a mapping project include:
1) A very large number of people likely don't live where the database says they do
2) Some sex offenders likely live in places they're not reported
3) More molestations occur by trusted people than strangers, e.g., the whole premise is misleading
4) Sex crimes aren't all standardized. I'm not terribly worried about a guy who got drunk and peed on a Dumpster 15 years ago, and thus got hit with an indecent exposure charge.
5) Some locations aren't where the geocoders say they are. I see 3 obvious errors out of about 700 entries in my sample database, though, so the error rate is likely better than I expected.
6) At present I have no way of showing multiple offenders in the same area. I see motels with a half-dozen guys, but if you click on a dot you'll see one. I found three convicted child molesters listed in the same motel *room*.

No guarantee of accuracy is implied, offered, donated, grilled, marinated or steamed. Don't sue me; you can take responsibility for verifying anything you do with this.


Mike Stucka
(478) 213-4742