Predictive Policing

[davidbfpo]

I have watched the developing debate around attempts to predict for a long time, which has become part of the so-called intelligence-led policing model and can be cited as a gain from using analysis / data-mining.

Taken from a paper by a US data-mining presentation:

X PD used data mining as a tool for determining how best to position police assets in anticipation of crime. Allocating assets so as to increase police presence where a particular incident is expected, for example, might help to prevent crime. She provided two examples of what she considered to be effective data analysis. One involved the application of supervised learning to the problem of random gun fire on New Year’s Eve. Data analysis was used to identify the times and places where the most incidents occurred. This information permitted local police to deploy officers strategically, resulting in a 47% reduction in the number of reported incidents and a reduction in personnel costs.

Link:http://www.detecter.eu/index.php?opt...&id=7&Itemid=9 within summary of the Zurich meeting on data mining.

One of the biggest issues around prediction and analysis is the data available, there is a considerable difference between actual / reported / recorded incidents and crimes. In the UK for example to officers dismay a large proportion of house burglaries are not reported. We have learnt, sometimes painfully, that low-level quality of life issues are far more important to the public than what the police want to do, such as "fighting crime".

[Majormarginal]

From your quoted article it seems that if the data mining predicts a crime and an arrest is made a crystal ball is successful. If nothing happens the crystal ball prevented a crime. Another success.

I have seen New Years Eve shooting cut down significantly from one year to the next. This reduction has to do with putting 300 police in a complex. This deployment does not happen until the swells move into the area.

[CWOT]

I've done a considerable amount of work on intelligence-led policing and I find this new thrust for predictive policing a real reach. I've yet to see any law enforcement agency conduct intelligence operations on the same scale or as continuously as the military-intel community. I think its great that police departments are finally reverting to intel processes and smarter employment, but I think most of their so called predictive policing techniques have been done by smart cops for years but not documented. Those that come from data are strictly extrapolations that don't really anticipate change as much as hope that future crime patterns match old crime patterns.

[davidbfpo]

The "think tank" Policy Exchange, who have considerable impact on UK government policies, have a seminar next week on 'Pre-Crime and Predictive Policing'.

From the summary:

Predicting where offences will occur and deploying police before crime happens has been an inexact science until recently, but that may soon change. Two pilots of the experimental ‘predictive policing’ method are underway in California and this new approach could have important lessons for UK policing in how forces deploy their resources to prevent crime....

The most robust predictive policing pilot, in Los Angeles, has just begun and shows some promising early results. The lead officer for the LAPD pilot, and a pioneer of predictive policing, Sean Malinowski, will be a speaker alongside George Tita, an expert on predictive models, from the University of California at Irvine.

Link:http://www.policyexchange.org.uk/eve...ent.cgi?id=405

(Added later in 2016) Link to video of the event:http://www.policyexchange.org.uk/mod...ctive-policing

[davidbfpo]

A lengthy article in The Guardian, sub-titled:

Smart technology and the sort of big data available to social networking sites are helping police target crime before it happens. But is this ethical?

He starts with:

The police have a very bright future ahead of them – and not just because they can now look up potential suspects on Google. As they embrace the latest technologies, their work is bound to become easier and more effective, raising thorny questions about privacy, civil liberties, and due process.

Link no longer works due to copyright and a search on Evgeny Morozov, the author's blog failed.

The author correctly draws attention to the leviathans of public use IT, for example Facebook & Amazon and asks who reviews their algorithms, for their ethical basis and effectiveness.

Given the clear failure to win the so called 'war on drugs', which has had massive funding and much hi-tech - why would this predictive policing be effective?

[Sergeant T]

The Economist weighs in on the topic.

Predicting and forestalling crime does not solve its root causes. Positioning police in hotspots discourages opportunistic wrongdoing, but may encourage other criminals to move to less likely areas. And while data-crunching may make it easier to identify high-risk offenders—about half of American states use some form of statistical analysis to decide when to parole prisoners—there is little that it can do to change their motivation.

I get a little queasy when we start handing decision making over to algorithms. As we used to say, you can't quantify the bad things that don't happen.

[davidbfpo]

Found this JHU author's paper whilst looking for something else, on a quick skim read it is a useful summary, but I still have my doubts over this approach.

Link:

http://www.businessofgovernment.org/...20Policing.pdf