SpotCrime is a mashup that plots recent criminal activity onto Google Maps, allowing users to shy away from seedier parts of towns they may not be familiar with. The company has offered a website for a few months now, and has just introduced an iPhone version (annoyingly, you have to manually navigate to http://www.spotcrime.com/iphone.php).

Crimes are depicted as small icons according to the type of incident, and users can filter crimes over a certain date range or time period. Clicking on an icon brings up more detailed information (when available). The site currently supports only a select number of (mostly large) cities, but it says that it is expanding quickly.

SpotCrime draws approximately 90% of its data from local police records, and, in cities where that information isn’t released, they gather it from local news sources. While most of this involves data scraping, SpotCrime says there there is still a human element required - something that will become increasingly problematic as the company grows. Incidents typically take 3-24 hours to show up in the system, so SpotCrime isn’t useful for avoiding crimes in progress. It does, however, provide a nice visual representation of criminal activity.

SpotCrime says that they are offering the service as a free tool to both police stations and the general population. While it’s hard to believe that the police don’t already know about the rougher areas of town, the visual overlays could conceivably held them identify trends. Other potential applications of the data include real estate evaluation, and (for more paranoid users) “safe” driving routes mapped by GPS.

SpotCrime isn’t the first site to do this - you can find similar offerings at The LA Times Homicide Map, UniversalHub, and a number of others. But most of these sites are regional, and SpotCrime is intent on cataloging cities across the country - something that will be hard to do unless they can find a way to fully automate their system.

Crunch Network: CrunchGear drool over the sexiest new gadgets and hardware.