Summary: these pages build up through a series of example techniques for analysing and predicting crime, introducing new elements to the analysis with each step. They start with relatively simple but powerful mapping techniques to look at crime patterns, and build towards sophisticated behavioural models.
CrimeStat:
We'll introduce briefly pattern revealing crime mapping and statistical mapping. This gives us the basic patterns of crime and therefore
patterns of risk, and allows us to identify long term hotspots of crime.
Risk Terrain Modelling:
We'll then look at Risk Terrain Modelling, a technique that looks for correlations between crime patterns and socio-environmental factors
to look at the causes and determinants of crime.
Near Repeat Modelling:
We'll follow this by looking at Near Repeat Modelling: a technique that uses distances in space and time between a series of crime events
that reveals the behaviour of criminals in a way that can then be used to make risk surfaces more accurate..
Microsimulation:
We'll then look Microsimulation, a methodology for estimating the demographics of individuals in an area and therefore building up
a population at risk..
Agent Based Modelling:
Finally we'll look at a technique that is in its infancy, but which promises to draw together understanding of socio-environmental determinants,
offender behaviour, the population at risk, and broader theoretical understanding into a framework that both allows us to assess our understanding of
crime but also potentially make in silico experiments that increase the accuracy of our crime predictions, both now and under conditions of
change. We will look at one specific example and talk about how to set out and
build your own ABM.