Slashdot Log In
Police Data-Mining Done Right
Posted by
Zonk
on Fri Aug 10, 2007 02:33 PM
from the way-its-supposed-to-be-used dept.
from the way-its-supposed-to-be-used dept.
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.'"
Related Stories
[+]
IT: FBI Data Mining For More Than Just Terrorists 130 comments
jcatcw writes "Computerworld reports that the FBI is using data mining programs to track more than just terrorists. The program's original focus was to identify potential terrorists, but additional patterns have been developed for identity theft rings, fraudulent housing transactions, Internet pharmacy fraud, automobile insurance fraud, and health-care-related fraud. From the article: 'In a statement, Sen. Patrick Leahy (D-Vt.), chairman of the Senate Judiciary Committee, said the report [on the data mining] was four months late and raised more questions than it answered. The report "demonstrates just how dramatically the Bush administration has expanded the use of [data mining] technology, often in secret, to collect and sift through Americans' most sensitive personal information," he said. At the same time, the report provides an "important and all-too-rare ray of sunshine on the department's data mining activities," Leahy said. It would give Congress a way to conduct "meaningful oversight" he said.'"
[+]
Schneier Talks to the Head of TSA 342 comments
Bruce Schneier recently had the chance to sit down with Kip Hawley, head of the Transportation Security Administration (TSA), and discuss some of the frustrations travelers experience head-on. "In April, Kip Hawley, the head of the Transportation Security Administration (TSA), invited me to Washington for a meeting. Despite some serious trepidation, I accepted. And it was a good meeting. Most of it was off the record, but he asked me how the TSA could overcome its negative image. I told him to be more transparent, and stop ducking the hard questions. He said that he wanted to do that. He did enjoy writing a guest blog post for Aviation Daily, but having a blog himself didn't work within the bureaucracy."
This discussion has been archived.
No new comments can be posted.
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
The Last Days of the Permanent Floating Riot Club (Score:3, Interesting)
A hole in space. [amazon.com] Larry Niven.
Are the police going to share the location information?
I might want to watch. Cops live!
Re:The Last Days of the Permanent Floating Riot Cl (Score:4, Insightful)
Strategic and tactical placement of resources to maximize effect without resorting to profiling or harassing citizens is a good thing.
Re:The Last Days of the Permanent Floating Riot Cl (Score:5, Insightful)
(Last Journal: Wednesday July 11, @08:27PM)
What about cops walking a beat? That's right, walking down the same streets over and over everyday. Walking a beat means getting to know the locals and the particulars of a neighborhood in a way that doesn't happen in a squad car. Gangs don't hang out on a corner if once an hour a cops walks by a says hello, but the neighborhood kids still can hangout and could even end up viewing that cop as part of their neighborhood. From a squad car, no relationship is established and any stationary pack of teens can look like a gang to a biased eye. You don't have to profile when you actually know the people you see, but if you are just cruising along looking at a sea of nameless faces, then ethnicity and clothing style are about all you have to go on. Profiling is almost inevitable without establishing officers with good personal knowledge/relationship with the locals.
Re:The Last Days of the Permanent Floating Riot Cl (Score:4, Interesting)
(http://whiskeyandtheweb.blogspot.com/)
What do the police say when they stop you? (Score:1, Redundant)
(http://www.videoproductionsupport.com/ | Last Journal: Thursday August 02, @01:49AM)
That's some fine police work, Lou. (Score:5, Funny)
Re:That's some fine police work, Lou. (Score:5, Insightful)
Re:That's some fine police work, Lou. (Score:5, Insightful)
(Last Journal: Sunday October 07, @01:01AM)
Still, I have to congratulate you. This just wouldn't be Slashdot if we didn't get somebody denigrating the accomplishment. It's very gratifying to know that I post to the same board as so many people who can do everything with merely a stray thought, if they ever actually felt like getting around to it.
Re:That's some fine police work, Lou. (Score:4, Insightful)
(http://penguin.lvcm.com/)
What happened to hunches and intuition?
The point of data mining to to find the NON-OBVIOUS relationships.
There's even a data warehousing product named just that.
Re:That's some fine police work, Lou. (Score:5, Interesting)
(http://www.burr.cc/~egburr)
Re:That's some fine police work, Lou. (Score:5, Insightful)
Whose payday? We can't post cops at all of the check cashing storefronts (not banks) all the time, so which stores see the most crime after which companies' paydays?
If you can answer these questions with nothing but hunches and intuition, I'd suggest the stock market, not law enforcement.
Re:That's some fine police work, Lou. (Score:5, Insightful)
(http://www.lazylightning.org/)
You know, when you get down to it, there's a lot of stuff that jumps out after the fact that says, "why the fuck didn't we notice that before?" But when you're doing the day to day work (in any field) you may ignore or not even know about what seems unbelievably obvious.
Just because this particular piece of information was the most prevalent in the story does not mean it's the only thing to come out of the reporting and it's certainly not the last thing that will be. Give it time. Data-mining's best fruits come from long term studies of data using a variety of methods.
Happens to us too (Score:4, Insightful)
(http://phorm.phormix.com/ | Last Journal: Monday May 19 2003, @12:08PM)
Sometimes it just helps to have somebody checking up on your work, even if that "somebody" is an automated process or machine.
Sweet! (Score:5, Funny)
(Last Journal: Wednesday February 14 2007, @01:57PM)
Which check-cashing place do you go to?
Interesting (Score:5, Interesting)
(http://thepeckfamily.us/ | Last Journal: Thursday November 08, @11:19AM)
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:Interesting (Score:4, Insightful)
(http://theari.com/)
Re:Interesting (Score:5, Insightful)
Says who? AFAICT, crime is best prevented by some minimum amount of personal freedom, reasonable living conditions regarding food, shelter, and education for all, along with some participation in matters of society.
Not really. (Score:5, Insightful)
Not really. Jail time and such has almost no effect on changing criminal behaviour.
Possibly. Or maybe they are trying to prevent crimes.
The criminals are not worried about going to jail AFTER the crime is committed. But if there is a cop there at the moment they would have committed the crime, most criminals will not commit it.
Means
Motive
Opportunity
With a cop right there, the "opportunity" is removed. So no crime occurs. In general, the crime rate should go down because this isn't something that can easily be displaced. It seems to be tied to the area around a check cashing storefront. Increase the patrols in those areas and the crimes are not committed.
A better way to prevent crime (Score:4, Insightful)
imho
Re:A better way to prevent crime (Score:4, Insightful)
(http://horsies.co.uk/)
TV (Score:2)
Wrong city listed in summary (Score:5, Informative)
Chip H.
How long until it catches on? (Score:2, Insightful)
(Last Journal: Thursday April 06 2006, @09:11AM)
Some people don't go to places at peak time to avoid queues, if criminals realise the police know the peak times, they can anticipate the strength of guard and where police are?
Knowledge like this can be used to both party's advantages. Some facts are obviously public knowledge such as weather.
I don't think it even takes well-organized crime to understand this.
How about the police force has a counter-itself division? It uses the public knowledge and works independent of the police to outsmart it -- the police can use this knowledge to anticipate counter knowledge usages...
Yes, but (Score:2)
Quick use the Drizzle Phone baby! (Score:1)
scanning the comments here on slashdot (Score:4, Insightful)
(http://circletimessquare.com/)
cop work is one of the most criticized, and yet at the same time vital, aspects of modern life
almost all the comments here have some sort of negative thought or smarmy remark on an aspect of this story. and yet a cop is the first person these same people will call upon and depend upon if they are ever victimized or robbed. and what are the cops doing? no, what are they actually doing? i'm not asking your paranoid distrustful hollywood-addled alter ego, i'm asking your cognitive ability to look at and perceive the reality of actual police work
typical human shortsightedness and lack of gratitude
it must be so thankless being a cop. you're there to protect people, and all they can do is reflexively depart negativity at you
humanity sucks. you are all so ungrateful
Re:scanning the comments here on slashdot (Score:4, Informative)
Re:scanning the comments here on slashdot (Score:5, Insightful)
(http://www.digitalsanctuary.com/)
Add that to MANY instances of being harassed by cops for my car, my youth, being out late, etc... and it's hard not to have a negative view. I'm sure there are some good cops out there. I don't doubt it. But when I'm robbed they can't send a cop out that day, presumably because they're all too busy pulling over young guys in expensive cars and searching them without probably cause (I'm in tech, I'm not a drug dealer), or issuing speeding tickets for 25 in a 20 to meet their quotas.
Anecdotes don't make a rule, it's true, but they do color a persons opinions. I've interacted with law enforcement many times (speeding tickets, random pull overs, having my apt. robbed, car accident, firearms testing for concealed carry, etc...), probably about 25 interactions. Of those, one was reasonably positive (helped after my car died on the side of the road), a few were neutral (neither helpful not malicious or abusive), and the rest (about 20) were negative (screaming and threats, searches without cause, rough handling, rudeness, apathy, etc...).
Re:scanning the comments here on slashdot (Score:5, Insightful)
(Last Journal: Wednesday October 05 2005, @10:39AM)
My point being: the police have different criteria for what's important than you do, and they're professionals with lots of experience. Your history with them sounds like it sucks, and it's likely they were wrong a lot of the time. But you don't know why they're doing what they're doing, and my observation is that their decisions don't seem to be completely arbitrary.
no kidding (Score:1)
What did this bit of information cost? (Score:1)
Well, duh!
Why? (Score:1)
Pre-Crime Unit (Score:2)
(http://www.burr.cc/~egburr)
That;'s what CompStat in NYC does (Score:3, Interesting)
(http://www.animats.com)
The NYPD's CompStat system [wikipedia.org] has been doing that for about ten years now. It's working reasonably well. At first it was really effective, because career criminals tend to fall into predictable patterns. Crime in NYC has dropped enough that there's more randomness, and prediction is less effective.
Hmmm (Score:1)
(http://www.enc0der.com/)
Predicting crime (Score:2)
I think this has been done before (Score:2)
(Last Journal: Saturday January 29 2005, @08:27AM)
Data mining? (Score:2)
(http://tirania.org/blog)