Research Review And Synthesis Of Predictive Policing Essay

1630 Words Apr 17th, 2016 null Page
Research Review and Synthesis
Algorithms and data are intrinsically biased. When a database is formed it contains the hidden assumptions and categories that people have, when a machine learning algorithm analyzes the database its strength lies in its abilities to acquire those hidden categories and act upon them[4,]. Acting upon hidden data is what the predictive algorithms are doing in predictive policing as well[6]. The tendency for one neighborhood to report more often than other automatically highlights that area to having more crime because of that tendency[1]. The algorithms ability to recognize implicit categorization is good most times but in scenarios such as predictive policing it results in biases becoming further entrenched in the algorithm. Seattle adopted predictive policing to distance itself from bias, when algorithms themselves may be encoding biases it signifies how necessary it is to reconsider the actions taken. Predictive policing and the actions taken by those in the system still need to be researched, reviewed and analyzed.
The existing research and analysis of predictive policing generally concludes that it helps. The results of the analysis say that it reduces crime, allows a greater efficient use of police resources, and that it is the answer to crime. The self reported numbers that PredPol reports to those who question its effectiveness say that it reduces crime[17]. In addition to reducing crime, predictive policing allows police departments to…

Related Documents