2013-07-22 – London
Can data analytics help reduce crime while protecting civil liberties and privacy?
New crime-prediction software is changing policing in Kent. The software which is used highlights likely crime hotspots and, according to experts in policing, it is surprisingly accurate. Kent Police’s trial of the software is just the latest example of police forces using big data to predict and prevent crime.
During a four-month trial in Kent, 8.5% of all street crime occurred within the predicted 150 square metre boxes, with a greater percentage occurring in adjacent areas. Predictions from police analysts scored only 5%. An earlier trial in Los Angeles saw software score 6% compared with analysts’ 3%. These may seem like modest gains, but they have excited police forces and industry experts alike, with police forces converting the modest prediction gains into significant crime reductions. Within six months of introducing predictive techniques in the Foothill area of Los Angeles in 2011, property crimes had fallen 12% compared with the previous year; in neighbouring districts they rose 0.5%.
Highly advanced reactive systems have been used by police forces for years; and although these systems have allowed the police to respond more quickly to incidents involving crimes that have already occurred, it has done little to prevent crime. New predictive policing, bringing with it a deeper analysis of data about past crimes combined with sophisticated algorithms, allow police to predict crime before it happens.
Privacy concerns
Whilst the benefits of predictive policing using big data are clear, so too are the privacy concerns. Given enough data and spurious algorithms, most individuals could end up looking suspicious. It is of utmost importance to develop highly accurate algorithms, and to protect private data, for predictive policing to work. This is especially true in free democratic societies which have strong civil liberties that need to be protected, according to Daniel de Bruin, managing partner of Modelling Design Partners, a data analytics consultancy. It is of no use to reduce crime, if individual liberties and freedoms are not protected at the same time.
Data analytics and data fusion
Predictive policing itself is not new. Police forces have been using crime analysts for well over 40 years. What is new is the fusion of big data from a multitude of sources. If privacy concerns can be assuaged, predictive policing looks set to revolutionise policing in the UK and around the world, allowing forces to reduce crime, save lives and improve public safety, whilst protecting individual freedoms and privacy.
