Follow Slashdot stories on Twitter

 



Forgot your password?
typodupeerror
×
Crime Technology

Police Program Aims to Pinpoint Those Most Likely to Commit Crimes 244

An anonymous reader writes: Using profiling algorithms, police are tracking suspected criminals to prevent them from committing predicted crimes. We're one step from locking people up for what they might do. The New York Times reports: "The strategy, known as predictive policing, combines elements of traditional policing, like increased attention to crime “hot spots” and close monitoring of recent parolees. But it often also uses other data, including information about friendships, social media activity and drug use, to identify “hot people” and aid the authorities in forecasting crime."
This discussion has been archived. No new comments can be posted.

Police Program Aims to Pinpoint Those Most Likely to Commit Crimes

Comments Filter:
  • Minority report. (Score:5, Informative)

    by sims 2 ( 994794 ) on Friday September 25, 2015 @11:54AM (#50597613)

    I've already seen this movie, And I think its a tv series now too.

  • by Anonymous Coward

    https://en.wikipedia.org/wiki/Minority_Report_(TV_series)

  • by JustAnotherOldGuy ( 4145623 ) on Friday September 25, 2015 @11:57AM (#50597653) Journal

    I cannot see anything that could possibly go wrong with this idea, except for everything.

    Can't wait until it's hacked and they start arresting the police chief, the city council, and everyone with a zip code that begins with a letter or a number.

    • funny how no one mentions that this is yet another government computer system that's vulnerable to hacks because the government's security is shit. Thanks for this, seems like everyone missed it. I would mod up if I had points left.
    • LOL - an even better tool for script kiddies than "SWATing".

      They'll be "MinorityReporting" people in no time!

    • Can't wait until it's hacked and they start arresting the police chief...

      Well, that would be nice if the cops pointed this program at themselves and root out those who are most likely to commit crimes against us. It would probably empty out the department though.

      • Can't wait until it's hacked and they start arresting the police chief...

        Well, that would be nice if the cops pointed this program at themselves and root out those who are most likely to commit crimes against us. It would probably empty out the department though.

        Easily fixed:

        if(subject_is_a_cop){
                  return 0;
        }

  • To quote Eddie Izzard, himself paraphrasing someone else:

    (running to one side of the stage)
    "I've got a new idea, I've got a new idea..."

    (turning, and running away in the other direction)
    "IT'S THE SAME IDEA! IT'S THE SAME IDEA!"

    The ability to predict crime has been the holy grail of law enforcement for over a century now. They've tried psychology, sociology, biology...even phrenology...to try and point the finger at people and say "Yep, that guy's gonna commit some crime; let's harass the living fuck out o

  • by MyLongNickName ( 822545 ) on Friday September 25, 2015 @12:01PM (#50597677) Journal

    Okay, folks, it is time to get rid of the political correctness and call a spade a spade.

    There is one group of Americans who cause the vast majority of violent crimes in the United States. They commit a disproportionately large amount of homicides and make up a vastly disproportionate amount of the people in prisons across the country.

    They are responsible for being a majority of drug offenders. They cause most fights in schools and make up the vast, vast majority of high school drop outs.

    They are also responsible for a vastly disproportionate amount of DUIs.

    But political correctness will not allow us to talk about this problem. If we simply locked up this part of society, we'd all be better off.

    And I think we all know what group this is.

    .
    .
    .
    .
    .
    .
    .

    Males. Lock 'em all up and throw away the key.

    • by sconeu ( 64226 ) on Friday September 25, 2015 @12:36PM (#50598113) Homepage Journal

      Let me guess... you're talking about politicians?

  • Selection bias (Score:5, Insightful)

    by DoofusOfDeath ( 636671 ) on Friday September 25, 2015 @12:02PM (#50597685)

    Their tool only works (if at all) on certain categories of crime. It misses huge ones, particularly those in which the people in power engage:

    * CIA torture.
    * CIA, NSA, FBI, and police violations of the Constitution.
    * CIA perjury to Congress.
    * Bank executives' racketeering in regards to mortgage origination.

    I'd pay good money for a tool to detect those kinds of crimes, and to see them properly prosecuted.

    • In the article one of the critics describes the general approach as:

      âoeBecause you live in a certain neighborhood or hang out with certain people, we are now going to be suspicious of you"

      So it's suspicion based on who you associate with. If you hang out with gang members, you might act like a gang member.

      Notice you called it "NSA, CIA, and FBI violations of the Constitution ". You didn't ascribe those crimes to isolated individuals. Instead, you're suggesting that many people involved the NSA have

    • I've developed my own algorithm for determining if these people are covering up crimes.
      It tracks lip movements, if there are any, they're covering up crimes.

    • by tnk1 ( 899206 )

      You may have a point, but you know, if law enforcement isn't having to bust as many small-time perps, maybe they will have some time to deal with the bigger fish.

      After all, we need cops to deal with domestic disturbances and robberies and such. If you could simply reassign some of the units busting those crimes to something else, you might see it affect the enforcement of other laws as well. Most cops aren't in the business to crack heads. They can get jumpy and shooty because they feel in danger or beco

  • I have no problem with monitoring parolees, that's part of the bargain. Wearing a tracking device to get out of jail early, submit to drug testing, etc. However, tracking people who have not committed any crime is a different story. Here's a scenario, a 'clean record' man loses his wife due to a drunk driver. He buys a gun and starts hanging out near the defendant (all realistically trackable by an algorithm collecting records & gps data), what is an appropriate response, if any?

    • He buys a gun and starts hanging out near the defendant (all realistically trackable by an algorithm collecting records & gps data), what is an appropriate response, if any?

      I think it is perfectly reasonable for the police, in such a situation, to talk to the person who is stalking the killer of his wife to advise him that he is going to be suspect number one if the guy winds up dead from a gunshot wound. I also think it is perfectly reasonable for the police to drive through the neighborhood on a regular basis, keeping an eye out for the "clean record" man's car and to let him know that they're watching.

      "Stalking" is a crime in many, if not most, places, and the threat is p

      • It seems in the last 5-10 years most people have started carrying a GPS receiver most places they go. Accessing this data currently requires a warrant, but is theoretically possible given the nature of the system being described.

        • Or just track their cell phones.

          • Or just track their cell phones.

            Yes, this is the "GPS receiver" that most people carry that the other poster was referring to. I carry two, typically, both not enabled.

            It required a warrant to get the cellular tracking data, or a signed statement that the data is needed for emergency (safety of life) purposes. And the latter only gets you the current location, not a track.

  • Great idea! (Score:5, Interesting)

    by dirk ( 87083 ) <dirk@one.net> on Friday September 25, 2015 @12:04PM (#50597697) Homepage

    I think this sounds like a wonderful idea. I propose as a test that we use it on police departments to try and determine those officers that are most likely to abuse citizens. If it is successful in dropping those number significantly then we can talk about maybe trying it out on citizens.

    • Great suggestion. Also, how about using it to find white collar criminal wannabees? Or better yet, predict which politicians will break the law.
  • by xxxJonBoyxxx ( 565205 ) on Friday September 25, 2015 @12:08PM (#50597749)

    If you've committed a crime, it's more likely that you, rather than someone who has never committed a crime, will commit the next crime. The term is "recidivism."

    If you've never committed a crime, I think it's about a 3% chance you'll commit a serious one. (http://www.naacp.org/pages/criminal-justice-fact-sheet) However, if you have committed a serious crime, you'll about 40% likely to commit another serious one within 3 years. (http://www.slate.com/articles/news_and_politics/politics/2014/09/recidivism_and_mental_illness_iowa_s_central_pharmacy_pilot_project_is_an.html)

    • by beakerMeep ( 716990 ) on Friday September 25, 2015 @12:28PM (#50597995)

      A lot of parolees are placed at a serioius disadvantage though.
      - Have trouble getting a job due to a record.
      - Often have to pay ridiculously high fines and "fees" related to their processing and incarceration
      - Rarely get good medical treatment for addictions.
      - Were often forced into the system by a plea deal

      If we are serious about preventing recidivism, we would lay better foundations for helping these people become productive members of society.

      Sadly, the reality is that our society prays upon those least able to defend themselves. No one wants to stand up for a convicted ______.

      Here's a link about the "fees" http://nation.time.com/2013/08... [time.com]

    • by swb ( 14022 )

      Some recidivism may be due to an inherent "criminal nature", but it sure seems like a good chunk of it is due to getting put back out on the street with a record that makes it impossible to get any kind of a decent job or stable housing, which in turn often forces offenders back into the same environment they came from, surrounded by the same people and situations that contributed (not caused) to their criminal activity to begin with.

      I don't know if it's true or not, but isn't murder one of those crimes tha

    • If you've committed a crime, it's more likely that you, rather than someone who has never committed a crime, will commit the next crime. The term is "recidivism."

      No, the term is "Puritanism", meaning once a criminal, always a criminal (especially if you ain't UMC white), with ZERO chance of going straight as the entire society is arrayed against you. You can't live (rentals only) in vast swaths of the country because you will only be ever hired (if at all) at the worst, lowest type of jobs reserved for t

  • This is good news for those profiting from the privatized Correctional system.
  • In this comment [slashdot.org]. Just wait until Watson is thrown into the prediction mix...
  • people spy on each other and report their neighbor, police spy on everyone: Soviet Union, Nazi Germany etc.

    USA is following a well worn path

  • Lets just cut the bullshit. Here we go, when a white male turns 18 he gets randomly sentenced from 10 to 20 years in prison. A black male gets 20 to 40 years and a Hispanic male gets 20 to 50.

    A women will get a random sentence from 5 to 10 years regardless of race.

    There that should do it. I think I covered police prejudices in my assessment quite nicely. Address all present and future crimes all in one setting

    if that doesn't work lets just send the police out to collect any random person they

  • I predict that the police will abuse this.
    perhaps they should use this tech on their own selves first.

  • IF (black) AND IF (shitty_neighborhood) AND IF (gold_crowns)
      THEN criminal = true
    ELSE
        criminal = false

  • If you catch them before they commit a crime based on a so-called prediction they would commit a crime, then they won't be able to commit the crime, which suggests that the prediction is invariably fallible. Either you predict that someone will commit a crime and stop them before they do making the prediction wrong, or else you fail to predict that someone will commit a crime in the first place. In both cases, the mechanism for predicting crimes fails unless it is not used to stop the crime from occurring
    • by robi5 ( 1261542 )

      Whoa whoa. Why is it that some people act rationally on a personal level, but lose reasoning power if the question is about the larger society.

      If stealth is going on in my dorm room, and I suspect someone, it's possible that I can catch him in the act, or after the fact. There are known techniques, like, you know, watching, or using bills with previously recorded serial number, or marking bills with UV ink. Similarly, for other types of crime, once you have a strong suspicion, it's possible to come up with

      • by mark-t ( 151149 )
        You can't charge someone with attempted murder until after they've already attempted it... In case you hadn't noticed, the tense of "attempted" is past tense. The most you can charge them with intent to commit murder, or conspiracy to commit murder.
        • You can't charge someone with attempted murder until after they've already attempted it...

          Wouldn't it be great if you could prevent the murder in the first place, by maybe letting the potential murderer know ahead of time that he's being watched? You know, like what this system is doing.

          And also let him know that other crimes will be dealt with harshly because, well, he's a multi-time offender and apparently needs more clues that he should stop than the first time offender needs? Some places have three-strikes laws. This is just letting people know they're getting close to the limit.

          I don't k

          • by mark-t ( 151149 )

            Of course it would be great to prevent a murder from happening in the first place, but that still doesn't mean you can charge them with attempted murder if they haven't actually already tried to do it. As I saId though, charging them with intent to commit murder, or conspiracy to commit murder would be entirely accurate.

            And any system that allegedly predicted they would murder someone in the future would in fact be proven by said thwarting to have been inaccurate, since the murder was prevented, it did

            • Of course it would be great to prevent a murder from happening in the first place, but that still doesn't mean you can charge them with attempted murder if they haven't actually already tried to do it.

              The program being discussed does not charge people with crimes they have not committed. Nobody says "we're predicting you have a high likelyhood of re-offending so we're going to arrest you now."

              And any system that allegedly predicted they would murder someone in the future would in fact be proven by said thwarting to have been inaccurate,

              No. Stopping someone from murdering someone else before they did it does not mean they would not have committed the murder. The fact that the system DID stop them shows that it was accurate, otherwise it would not have stopped them. (You cannot "stop" someone from doing something they weren't going to do. If you

              • by mark-t ( 151149 )

                It's not faulty logic... it's a reformulation of the classic halting problem, a well known paradox in computer science, and founded entirely on very solid reasoning.

                My point is only that it either fails to predict a murder would happen because the murder is thwarted before it can, it fails to predict a murder in the first place, or else it predicts a murder that nobody will be able to stop. In the latter case, it's worthless, and the other two cases, it's wrong.

                That a system might be developed that c

                • It's not faulty logic... it's a reformulation of the classic halting problem, a well known paradox in computer science, and founded entirely on very solid reasoning.

                  The halting problem does not include the concept of "I predict it will not halt. I have, however, chosen to manually reset the computer, thus halting the program artificially, and therefore the prediction that it would not halt has been proven incorrect." That's what you are doing if you chose to detain or prevent the murder you have predicted will occur. You have not proven the prediction to be wrong, you have acted outside the system to force an outcome that the original prediction system did not conside

        • You can't charge someone with attempted murder until after they've already attempted it... In case you hadn't noticed, the tense of "attempted" is past tense. The most you can charge them with intent to commit murder, or conspiracy to commit murder.

          Conspiracy to murder is an extremely serious offence, in the UK at least it carries a potential life sentence, the same as murder.

  • In the US, crime has become pretty much arbitrary in definition and enforcement. So this will be very bad, and it will breed massive crimes.
  • How about a tool to predict which officers are most likely to commit unjustified shootings of civilians?

    Then we can fire them pre-emptively BEFORE they murder innocent people!

  • The next time they, or anyone in their crews, commit a violent act, the police will come after everyone in the group for whatever offense they can make stick, no matter how petty.

    So... minor offenses that good white people can get away with will now be persecuted (with zeal) against minorities mostly. Nothing there that will inflame some community like Ferguson or Baltimore.

  • What a lot of words to say "we're going to continue harassing the poor and minorities".

  • From TFA:

    "The goal is to do all they can to prevent the crime from happening."

    How can you know if you have succeeded? How do you know what would have happened if you hadn't done something? Unfortunately, reality does not have a control group.

    • "The goal is to do all they can to prevent the crime from happening."

      How can you know if you have succeeded?

      In a single case, ultimately, if want to be really really pedantic about i? You don't. I mean, that kid might have had a perfectly non-lethal reason for bringing a loaded shotgun to school, but that's not going to cause the cop who arrests him before he can (potentially) begin a rampage to lose any sleep.

      So no, if you want to get all deep and meaningful and armchair-philosophical about it, if you succeed you won't know, for sure. But that's a lot better than failing by not bothering to try.

    • From TFA:

      "The goal is to do all they can to prevent the crime from happening."

      How can you know if you have succeeded? How do you know what would have happened if you hadn't done something? Unfortunately, reality does not have a control group.

      You would obviously have to look at over all crime statistics. I know this will horrify the rugged individualists and libertarians here.

  • //Algorythm

    if not race="caucasian"
        possible_perp=true
    endif

    • //Algorythm

      if not race="caucasian" possible_perp=true endif

      The problem with you and other people who says stuff like this is that they don't think they are racist.

  • If they're a drug user, poor, or black they're multiple times more likely to commit a crime. Unfortunately you can't live in reality or use math if you're a cop. You have to pretend everyone is equal and waste time and resources accordingly.
  • Crime and poverty go hand in hand. Usually poverty occurs among minorities more than upon the majority race. Better said crime that is easy to detect is common among minorities. Crimes behind mansion walls are normally not in the eye of cops. Predictive policing can be as simple as recording what area has the most incidents at a certain time of day and then putting extra cops at that location at that time of day. The effect can be racist simply because cops are evaluated from the number o
  • Is it better to assign much more police effort to white collar crime or to arrest 1,000 people trying to buy crack or carrying an illegal weapon?

    The police and the DA will always go after low hanging fruit as this makes them look good when they tout statistics on how "tough on crime" they are.

  • We need to start sending copies of the U.S. Constitution out. It seem our lawmaker don't know a f'ing thing about it.

  • California (or at least in LA county) already has increased penalties for criminal association (ie, gang membership). This isn't RICO-type penalties. These don't criminalize advise-and/or-coordination of criminal activities. They simply add jail time, if convicted, to acts already recognized as criminal acts. In other words, having friends who are gang members could potentially be used against a person to increase their penalty (years in jail) for non-coordinated criminal activity. This is not in Texas

  • There's a lot of milage in engaging those most likely to commit crime and leading them away from that path.

    The US fixation on "revenge" rather than "justice" is rather worrying.

What is research but a blind date with knowledge? -- Will Harvey

Working...