Back in December, the city of Philadelphia published its first really major municipal dataset, a record of all “Part I” crimes in the city (these are the really bad ones: homicides, rapes, robberies, aggravated assaults and thefts) dating back to Jan. 1, 2006. Each incident is tagged to a location in the city, making it possible to map assaults across the years, or vehicle thefts near your home address within the last 30 days.
Not surprisingly, in the short time since the city released this trove, a number of developers have already repackaged it. Nearly every week, says Mark Headd, Philadelphia’s chief data officer, the city has stumbled across some new application of the material created by local residents. “It’s really interesting when you put valuable data out there, data that resonates with people – and this clearly does, and we make it really, really easy to use," Headd says. "It’s amazing what pops up."
The city itself initially mapped the data here:
But the really impressive applications are coming from outside of City Hall. This is the PHL Crime Mapper, created by software developer David Walk. His tool enables users to draw a polygon around any area in the city. PHL Crime Mapper then maps crimes within that territory during a given time span (and spits out some relevant statistics):
Detailed information on the suspect arrested in any given crime then links to another project of McGill’s, the Philly Rap Sheet, which scans the city’s court systems every half hour for data on the arresting officer, assigned judge, charges and bail for each crime. Here is the corresponding Rap Sheet page for the above assault:
Headd also created a visualization of his own, an animation of homicides across the city since 2006 (below is just a static screen grab, but you should click here for the full effect of the time series):
All of these tools have taught the city, Headd says, that municipal data is most valuable when people can parse it down to the level of their own communities. “In a city like Philadelphia, the story on crime can be starkly different neighborhood to neighborhood,” Headd says. “People want to be able to ask their own questions, or present their own take on this data.”
And that lesson, he adds, will likely inform how the city rolls out its next big datasets (look next for some financial and budget data). When a city releases data like this, an interesting thing also happens beyond its borders. That homicide time series Headd recently built on an open platform? It’s already been copied in Oakland. Here’s a video, from Steve Spiker, illustrating 18 years of homicides in two minutes: