Dr. Rick Adderley: Abstract

In this current economic climate all UK Police Forces have been required to trim their budgeting requirements which means that they will be losing staff. Less Police Officers and Police Staff to perform the same amount of, or even more, daily tasks.

On 14th July 2011 the Independent (and other papers) reported an increase of 14% in burglary offences, opportunistic thefts have also increased. How can this problem be addressed? By using a data mining tool that permits the development of automated of intelligence data processing one large UK Force has stated that it can save £250,000 annually be adopting just one single process that identifies its priority and prolific offenders in each of its areas. By saving that amount, their return on investment can be realised in less than one month.

A further process that uses a multi-layer perceptron can model personal behaviour and suggest which un solved burglary crimes may have been committed by an individual offender. This process is up to 85% accurate and takes about 10 minutes, traditional methods are typically 15% accurate and can take up to two hours to complete.

A further process using a self-organising map makes it possible to identify emerging crime series and even suggest one or more offenders who may be responsible for those crimes. Therefore, by adopting similar data mining techniques it can be possible to maintain or yet improve the service delivery of a Force even with a reduction of staff.

Go back

Organised by BCS SGAI The Specialist Group on Artificial Intelligence http://www.bcs-sgai.org