Friday, March 15, 2013

Can Big Data Reduce Gun Violence Ethically?


Can Big Data Reduce Gun Violence Ethically?

            Gun violence in this country is a huge topic in recent times. With school shootings and theater shootings the media is creating a feeding frenzy for anti-gun advocates resulting in a run on guns, ammunition, and magazines by gun advocates.  It seems though no one wants to find the root cause of the issue. The blame typically is placed solely on the instrument being used. But, can we as students and scholars find other underlying causes of gun violence, using big data analysis, or should we allow the gun to be the scapegoat that the media needs, in order to increase their ratings? I think Big Data can be used to uncover areas of interest that could be utilized to reduce the gun violence in this nation without limiting anyone’s 2nd amendment right or invading their privacy.    

            Data for violent crimes and gun related crimes is available in a limited capacity via the FBI’s Uniform Crime Reports (UCR) website. This data is already condensed and summarized for the most part so only limited knowledge can be gained from its usage. But, that does not mean it is not a useful source with the ability to point you down the right path to some underlying issues to gun violence. What we need is raw data that can be mined with an open mind and zero agenda clouding our judgment. We must establish where violent crime is committed and any associations it may have. This would be a similar to the tactics used in data mining and word association. I believe the answers we need lie in analyzing the past events and constructing solid associations. Asking simple questions like why did this happen? What demographics are more prone to participate in violent crimes? Once we know who the problem is let’s fix it! That maybe focusing on improving their education or the communities they live in.        

            Yet again there is a huge hurtle we scholars face, the ethical collection of data. In the United States we have many laws regarding privacy and we can’t go haphazardly gathering data. If we do, we are bound to cross the line and gather data that is an invasion of one’s privacy. In reference to Youngchul Yang’s blog linked below, creating a nationwide gun registry could very well be walking that slippery slope to invasion of privacy. Also using that data to decide who is more likely to commit a crime, that they haven’t committed, is a disaster waiting to happen. The problem with making broad assumptions on this type of data is allowing judgment to be passed without a crime being committed. This could be a very real threat in the future if the collection of data used in big data analysis isn’t monitored carefully.

                I think big data has its place in the reduction of gun violence, but we need to get more students and scholars involved and thinking outside the box to find these root causes. Not judge people by their buying habits or their hobbies and mark them as possible suspects. That is the easy way out! Resolve the issue at its source. You extinguish a fire by spraying its base not its flames…         

Sources:


1 comment:

  1. I agree with Julian in that ethics is the largest hurdle in using big data to detect criminal behavior. While privacy and gun ownership are not viewed as important in some countries, these subjects are often in the forefront of American politics. So when trying to predict gun violence, the questions that need to be asked are “What data is relevant/necessary?” and “If we can’t get all the data we need, are there other predictors that work just as well?”
    Would a national gun registry work? The FBI statistics provided by Julian’s link shows that Washington D.C. has the highest violent crime rate in the country despite having a gun registry and a strict gun ban. Even if a registry was implemented, how effective would it be in determining who will harm someone? Its effectiveness would be determined on how willing the population is to come forth and register their weapons, and the assumption that criminals use registered weapons. This harkens back to the importance most Americans place on privacy. Judging by the reaction of some New York residents about the state’s new gun laws, that might be a difficult proposition.
    Also as Julian stated, the key to predicting violent crime is determining which data to use in the prediction of violent crime. Using gun ownership might be minor factor but I doubt it would be deemed significant in a regression analysis since the fraction of gun owners linked to violent crimes. The following link the number of violent crimes/ 100,000 residents and the percentage of households in each state that posses a loaded firearm.

    http://www.datamasher.org/mash-ups/crime-vs-gun-ownership?page=1#table-tab

    The main thing to remember in predicting gun violence is the fact that humans are not machines. When it comes to certain behavior, violence in particular, human emotion plays a huge role. Emotions are such a wild card that researchers are still trying to classify emotions that motivate violence into distinct variables. To further compound the problem, human emotions are heavily influenced by past experience, genetics, and group dynamics. For instance, disgust and anger are very closely related emotions. The FBI proposes that “Elimination Based on Disgust” is the finally phase for premeditated violence. Most of us get angry, but rarely commit a violent crime. Anger can be a force for positive change, however disgust is rarely constructive. The key is to detect physiological differences in these emotions.
    Sources
    http://www.thenewamerican.com/usnews/constitution/item/14322-gun-owners-refuse-to-register-under-new-york-law
    http://www.fbi.gov/stats-services/publications/law-enforcement-bulletin/january-2012/the-role-of-emotion-in-predicting-violence

    ReplyDelete