2016년 5월 29일 일요일

Left Hand, Right Hand


Left Hand, Right Hand
--Data science is helping law enforcement
by Jia Wang, Minseok Song

In 2005’s report of the European Union research, European research suggests, “increased use of crime prevention measures may indeed be the common factor behind the near universal decrease in overall levels of crime in the Western world”(Wikipedia). Data science is acting as a useful hand to improve policing operation and enforcement. In this session, we will discuss about who are the active users of data science in law enforcement, what the problems they are addressing and what the challenges they are facing.

WHO

When seeing the power of data science, the law enforcement departments who have collected piles of data over the years are keen to grasp this useful hand to help them enforce the law. Most western government and police departments are very active dealing with this. With the alarm of IS attack, the European Union is also trying to apply big data technology to analysis airline passenger data to track any more terrorist crime and prevent heartbreaking tragedy (Europe, 2016).

Collaborating with government departments, vendors and developers are providing predictive tools and financial tool. These vendors and developers are usually IT companies or university research groups. As one of the world largest IT companies, IBM starts at a very early stage to help law enforcement institutions to fight with crime. Many world well-known universities such as university of Michigan, Georgetown University, MIT and so on are also enthusiasm in providing law enforcement solutions.

Problems

With the help of data science, law enforcement departments are mainly trying to address two problems: vaccinate crime and prevent official violent enforcement.

As we mentioned last week, one of the data science in law enforcement fronts is predictive policing. Governments try to reduce crime by collecting data, analyzing data by using predictive tools and operating officially. The effect is obvious. One example from an article is that the crime reduced 15%-30% in Santa Cruz, Calif. after they implemented a predictive policing tool called PredPol (Alexis C. Madrigal, 2016). Through the approach, predictive policing has been formed a valuable cycle as follows:


Predictive policing business/process cycle

The governments are also trying to avoid unnecessary dead encounters made by police officers. In America, the white house lunched the Police Data Initiative (Megan Smith, Roy L.Austin, JR., 2015) project to prevent violent enforcement of police officers. About 21 state police departments have participated in this project, mainly from the coastal regions, such as PA, NC and CA. The Police Data Initiative is particularly aimed to address officers who may behave violently and take improper action towards crime under certain situations. Supervisors will get warning in advance by the intervention system and take actions on these officers.


Police Data Initiative participants distribution

Challenges

Although a high technology the data science is, it could not be perfect. Followings are some challenges that data science in law enforcement may face:

1.     Narrow predicted crime area. It is easy to find probable hot spots with big range of geography, but hard to narrow the police operation in a small particular area.
2.     Data Quality. When dealing with multiple systems, aggregating data from these systems may be challenging. Also, data may be biased. Some criminal actions may not happen exactly on what it is reported due to some lag of witness time (Clarence Wardell, 2015).
3.     Budget. No matter doing what kind of projects, financial investment is important. Policing is big in certain situation. Financial is hard to limit.
4.     Privacy. Some crime data involves victims and teenager commitments. Sharing these kind of data with public in some cases (for example, share with university groups) is really an issue of respect and privacy rights.


Although data science in law enforcement is facing challenges, the effect of that in crime reduction and regulate police officers is significant. Data science is the left hand and policing operation is the right hand. They are bundled together to make the world.

References

Larry Greenemeier.(2015, 7 22). Can Police Use Data Science to Prevent Deadly Encounters? Retrieved 5 27, 2016, from Scientific American: http://www.scientificamerican.com/article/can-police-use-data-science-to-prevent-deadly-encounters/

Jennifer Markert.(2015, 3 30). Predictive Policing: Could Infusing Law Enforcement With Data Science Stop Crime In Its Tracks?Retrieved 5 27, 2016, from Curiousmatic: https://curiousmatic.com/predictive-policing-could-infusing-law-enforcement-with-data-science-stop-crime-in-its-tracks/

Walter L. Perry, Brian McInnis, Carter C. Price, Susan C. Smith, John S. Hollywood
. Predictive Policing. Retrieved 5 27, 2016, from: http://www.rand.org/content/dam/rand/pubs/research_reports/RR200/RR233/RAND_RR233.sum.pdf

Alexis C. Madrigal.(2016, 3 28).The Future Of Crime-Fighting Or The Future Of Racial Profiling?: Inside The Effects Of Predictive Policing. Retrieved 5 27, 2016, from TheWorldPost: http://www.huffingtonpost.com/entry/predictive-policing-video_us_56f898c9e4b0a372181a42ef


Clarence Wardell.(2015, 5 25).OpenGov Voices: Challenges and solutions to collecting law enforcement data.Retrieved 5 27,2016, from Sunlight Foundation: https://sunlightfoundation.com/blog/2015/05/25/opengov-voices-challenges-and-solutions-to-collecting-law-enforcement-data/

Megan Smith, Roy L.Austin, JR.(2015, 5 18).Launching the Police Data Initiative.Retrieved 5 27,2016, from the White House: https://www.whitehouse.gov/blog/2015/05/18/launching-police-data-initiative

Europe.(2016, 4 13).Passenger Name Record: EU to harvest more data to stop crime

.Retrieved 5 27,2016, from BBC News: http://www.bbc.com/news/world-europe-36035698

Wikipedia
https://en.wikipedia.org/wiki/Main_Page




댓글 없음:

댓글 쓰기