By CRAIG BYRNES
Amazon.com employs analytical techniques in each and every facet of the consumer experience. Their predictive algorithms bring together association rules, cluster modeling, and a vast array of consumer usage data to bring each customer an experience meant to be uniquely theirs. Alternatively, it comes together to ease the shopper experience. These same analytical techniques can be adapted in order to serve law enforcement at every level, but most notably by taking advantage of the same techniques in regard to social media analytics.
Social media analytics can pick out patterns and information paramount to the security of a community, an event or a public figure. This process is already in place in Medford, Oregon where gangs have begun to utilize social media for their recruitment and retention efforts (Cauguiran, 2012). Police officers monitor postings based on keywords on Facebook, Twitter, YouTube and the like, and solidify their efforts with targeted opinion, network and text mining. As a result valuable intelligence is gained and the otherwise unquantifiable is quantified.
This same concept was demonstrated in the finance sector by Capital One (Davenport & Harris, 2007). This company decided to compete on analytics with the aim of getting to know who their ideal customer would be. By discovering they were able to target their market thus resulting in heightened service. Careful attention to detail in determining classifications and global attributes required for gaming future scenarios make this technology and subsequent fact-based analysis more likely to succeed in mitigating the number of decisions and predictions based on intuition theoretically resulting in a more accurate forecasting an ideal candidate for committing a crime or an ideal sector where a crime may be committed.
A recent article written by Dale Peet on the subject points out that this capability, “is still new territory to law enforcement and emergency management personnel,” but goes on to say, “it can’t remain new for long,” (City, 2012.) A plethora of data exists in open source to which analytical tools such as sentiment analysis can be applied. This is an intrinsically useful analysis for an array of businesses like Amazon.com and Wal-Mart, but is equally of worth to law enforcement. That said, it is important to note that actively scrutinizing social media for likely instances of ‘e-Thugging’ is not the only advantageous feature of this analytical tool (Cauguiran, 2012).
“Social media analysis tools, such as sentiment analysis technology, also can be useful in cases where threats have been made against public figures and other people,” (City, 2012). Peet explains that in January 2011 Jared Loughner, the person responsible for shooting Arizona Congresswoman Gabrielle Giffords, posted anti-government rants on social media sites before the assassination attempt,” (City, 2012). It is a stretch to presume preventative interdiction would have been possible simply as a result of these postings being scooped up in a data mining operation, but it is a step in the right direction setting the groundwork for prediction.
Law enforcement can also make sizeable steps toward a safer society by applying the same analytic techniques used by Amazon.com usually tied to areas of market research. Swarm analysis and swarm-based opinion prediction can potentially lead to the development of realistically conceivable preventative strategies of crime prevention leading to enhanced public safety (Bodendorf, 2011). This method should yield early warning of situations deemed critical to public safety or security by putting in place a systematic process of data collection governed by proper classifications and coding (Bodendorf, 2011).
What emerges is a more intelligible picture of relationships that bring patterns to light yielding trends that can be monitored for anomalies. By thinking outside normal parameters, and putting into place data mining techniques based on predictive and customized algorithms a safer community can be obtained.
Note: There will always, “be a trade-off between privacy and utility or, in other words, between disclosure risk and information loss,” when dealing with data mining activities and information sharing across the Internet be it open source or not. It will be interesting to see how the line progresses as the technologies and capabilities evolve (Torra, 2012).
Bodendorf, P. D. (2011, January 23-28). Social Media Analytics. Retrieved from GlobeNet 2011: http://www.iaria.org/conferences2011/filesDBKDA11/Globenet11_Keynote_FreimutBodendorf.pdf
Cauguiran, C. (2012, March 26). Gang activity is popping up on social media. Medford, Oregon, United States.
City, A., & Contributor, C. (2012). Viewpoint: To aid emergency management, turn to social media. The American City & County, , n/a. Retrieved from http://search.proquest.com/docview/1038035291?accountid=8289
Davenport, T. H., & Harris, J. G. (2007). Competing on Analytics: The New Science of Winning. United States of America: Harvard Business School of Publishing Corporation.
Torra, G. N.-A. (2012). Privacy-preserving data-mining through micro-aggregation for web-based e-commerce. Bellaterra, Spain: Artificial Intelligence Research Institute.
Related articles across the web