Modus Operandi Modelling of Group Offending: A Data-Mining Case Study

Abstract

This article looks at the application of data-mining techniques, principally the multi-layer perceptron, radial basis function and self-organising map, to the recognition of burglary offences committed by a network of offenders. The aim is to suggest a list of currently undetected crimes that may be attributed to one or more members of the network and improve on the time taken to complete the task manually and the relevancy of the list of crimes. The data were drawn from four years of burglary offences committed within an area of the West Midlands police. They were encoded from text by a small team of specialists working to a well-defined protocol and analysed using the above techniques contained within the data-mining workbench of SPSS/Clementine. Within minutes, three months of undetected crimes were analysed through the Clementine stream, producing a list of offences that might be attributed to the network of offenders. The results were analysed by two police sergeants not associated with the development process who determined that 85 per cent of the nominated crimes could be attributed to the network of offenders. To produce a manual list would take between one-and-a-half and two hours and be between 5 per cent and 10 per cent accurate.