--HELP / ADVICE--
I am a research student recently started at the University of Glamorgan
investigating techniques and methodologies for the spatial analysis of
violent crime utilising GIS as a tool. I am attempting to embrace questions
such as Where? When? Why? (and What can be done to reduce the problem?)
Background: Crime rates differ between different types of urban district,
and these differences are best explained by the variation in use of urban
sites by differing populations. Whilst a database of violent incidents
(location, cause, victim details including injuries and address) is rich in
spatial information, studies to date have been limited to simple statistical
analyses of these variables. However, a much richer survey can be undertaken
by linking this database with other spatial databases, such as the Census of
Population and police databases. Such information can tell us a lot more
about the environment in which the incidents take place; and the home
environment of the victim and/or perpetrator. Whilst geographical
information systems (GIS) can be used to identify simple trends, i.e. the
relationship between unemployment and violence at ward, enumeration
district, or postcode level, other not so obvious trends may lie
undiscovered in the database. Data mining techniques have recently gained
attention as a means of discovering such hidden information within
databases. One of the more popular approaches to data mining utilises neural
networks to detect trends.
The project will entail implementing a prototype GIS which integrates the
various spatial databases discussed above to analyse and map the identified
trends. The analytical engine will be supplemented with a neural network for
some of the processing.
I am essentially in search of any information that anyone feels may be
pertinent to the nature of this research and would be grateful for any
advise or comments.
Keywords: Crime Analysis, Geographical information systems, GIS, data
mining, census of population, spatial analysis, neural networks.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|