Use this URL to cite or link to this record in EThOS: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.763228
Title: Space-time modelling of terrorism and counter-terrorism
Author: Tench, Stephen Ashley
ISNI:       0000 0004 7660 728X
Awarding Body: UCL (University College London)
Current Institution: University College London (University of London)
Date of Award: 2018
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
Abstract:
In this thesis multiple approaches are presented which demonstrate the effectiveness of mathematical modelling to the study of terrorism and counter-terrorism strategies. In particular, theories of crime science are quantified to obtain objective outcomes. The layout of the research findings is in four parts. The first model studied is a Hawkes point process. This model describes events where past occurrence can lead to an increase in future events. In the context of this thesis a point process is used to capture dependence among terrorist attacks committed by the Provisional Irish Republican Army (PIRA) during ``The Troubles'' in Northern Ireland. The Hawkes process is adapted to produce a method capable of determining quantitatively temporally distinct phases within the PIRA movement. Expanding on the Hawkes model the next area of research introduces a time-varying background rate. In particular, using the Fast Fourier Transform a sinusoidal background rate is derived. This model then enables a study of seasonal trends in the attack profile of the Al Shabaab (AS) group. To study the spatial dynamics of terrorist activity a Dirichlet Process Mixture (DPM) model is examined. The DPM is used in a novel setting by considering the influence of improvised explosive device (IED) factory closures on PIRA attacks. The final research area studied in this thesis is data collection methods. An information retrieval (IR) tool is designed which can automatically obtain terrorist event details. Machine learning techniques are used to compare this IR data to a manually collected dataset. Future research ideas are introduced for each of the topics covered in this dissertation.
Supervisor: Fry, H. ; Gill, P. Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.763228  DOI: Not available
Share: