Police Data-Mining Done Right

enharmonix writes "Courtesy of Bruce Schneier, it's nice to hear something good about data mining for a change: predicting and stopping crime. For example, police in Redmond, VA, 'started overlaying crime reports with other data, such as weather, traffic, sports events and paydays for large employers. The data was analyzed three times a day and something interesting emerged: Robberies spiked on paydays near cheque cashing storefronts in specific neighbourhoods. Other clusters also became apparent, and pretty soon police were deploying resources in advance and predicting where crime was most likely to occur.'"

Comments

Interesting

[stoolpigeon]

I don't really tend to think in terms of the police having the job of preventing crime. I think there job should be to apprehend criminals who are involved in or have committed a crime. That said, I guess it is good if they have tools that better help them to schedule and plan enforcement. Like anything, it can be taken too far. I would think that what would separate 'good' data mining from 'bad' data mining would be transparency and over site in the process.
 
On a side note - I'm willing to bet that if someone had asked most street cops in that area - they wouldn't have needed software or data mining tools to tell you that cash checking places in bad parts of town, on pay days were areas of higher crime.

Re:That's some fine police work, Lou.

[egburr]

They probably are, but they can't admit it, because without hard data to back it up they get criticized for "profiling".