Thesis title:
Spatial Dynamical Networks of Concentration & Dispersion of Crime (a joint ESRC/EPSRC-funded studentship)
Abstract:
Studies in the geography of crime confirm that over 50% of crime incidents occur within 5% of the streets in each city, thus forming micro-crime places known as crime hotspots. For this reason, hotspot policing has generated considerable interest, whereby identifying high-risk places by means of hot spot detection and targeting hot spot locations in the practical policing scenes have come to form a pillar in efforts to reduce crime. A method recently proposed for crime hotspot detection is effective in identifying the exact times and locations of micro-crime places. However, it is retrospective in nature with its main utility being the promotion of long-term policing by reducing opportunities for crime. In other words, they do not necessarily help predict where and when crime will likely happen in the future.The aim of this project is to establish a new type of geosurveillance method for predicting the likely elevation of crime activities at the disaggregate street address level. The project involves the use of syndromic surveillance, originally designed for epidemiology and subsequently repurposed and refined in the crime analysis context to be taken to the next level by incorporating hidden Markov models. Nicholas will work closely with his supervisors to (1) explore how these two methods can be integrated into one, (2) code the new method as a workable tool, and (3) carry out empirical analysis using real-world crime data so as to confirm the validity and the performance of this method for detecting and predicting crime hotspots. Crime data will be provided by the Metropolitan Police Service.
First supervisor:
Naru Shiode
Pathway:
8 – Urbanisation, Social Change & Urban Transformation
Cohort:
2017-18