Predictive Policing & Its Usage

Predictive policing is the application of analytical techniques to identify potential targets for police intervention and solve crimes by applying statistical techniques.
Countries like UK and US see the use of predictive policing. Another term used to describe the analytical technique is called forecasting.

Predictive policing is the practice of identifying the date, time and location where specific crimes can occur which is followed by scheduling officers in those area in hope that the crime does not take place, thus keeping the neighbourhood safer.
Predictive analytical models are continuously refined with the help of software suppliers.

It can be categorised into four types of methods –
Predicting crimes – It is used to forecast places and time which are vulnerable to crime

Predicting offenders – This approach helps identify individuals who are at a risk of offending in the future.

Predicting Perpetrators identities – These techniques are used to create profiles which match offenders with specific crimes committed in the past

Predicting victims of crimes – They are used to identify groups or individuals who are likely to become victims of crime.

Usage of Predictive Policing

Predictive policing needs to be used part of a comprehensive prevention strategy which should add value to the police work. It can avoid certain pitfalls which include –

Focusing on predictive accuracy rather than tactical utility – Identifying the whole city as hot spot may be accurate, but it does not add to the information a police officer have. To ensure predicted hot spots are small and actionable, a limit on accuracy needs to be measured as a proportion to the crimes occurring in such hot spots.

Poor quality data – Data may suffer due to omission, how it is collected and relevance. It may be omitted for certain incidents, the collection of data may vary as well as in some cases collecting data from the past may be useful while on the other hand some cases may not require such old data.

Misunderstanding the factors – Observers tasked with identifying a hot spot, may not be able to adequately answer about a given hot spot and what factors are driving risk. When applying techniques such as data mining, using common sense to vet the factors can help avoid variations among factors.

Civil liberty – The act of labelling certain areas which requires increased law enforcement raises concerns about privacy rights. Labelling areas as ‘at-risk’ appears to have fewer problems as in that case individuals may not be targeted directly.