Please use this identifier to cite or link to this item: http://theses-test.ncl.ac.uk:8080/jspui/handle/10443.1/3919
Title: Tools and techniques for analysing the impact of information security
Authors: Mace, John Charles
Issue Date: 2017
Publisher: Newcastle University
Abstract: The discipline of information security is employed by organisations to protect the confidentiality, integrity and availability of information, often communicated in the form of information security policies. A policy expresses rules, constraints and procedures to guard against adversarial threats and reduce risk by instigating desired and secure behaviour of those people interacting with information legitimately. To keep aligned with a dynamic threat landscape, evolving business requirements, regulation updates, and new technologies a policy must undergo periodic review and change. Chief Information Security Officers (CISOs) are the main decision makers on information security policies within an organisation. Making informed policy modifications involves analysing and therefore predicting the impact of those changes on the success rate of business processes often expressed as workflows. Security brings an added burden to completing a workflow. Adding a new security constraint may reduce success rate or even eliminate it if a workflow is always forced to terminate early. This can increase the chances of employees bypassing or violating a security policy. Removing an existing security constraint may increase success rate but may may also increase the risk to security. A lack of suitably aimed impact analysis tools and methodologies for CISOs means impact analysis is currently a somewhat manual and ambiguous procedure. Analysis can be overwhelming, time consuming, error prone, and yield unclear results, especially when workflows are complex, have a large workforce, and diverse security requirements. This thesis considers the provision of tools and more formal techniques specific to CISOs to help them analyse the impact modifying a security policy has on the success rate of a workflow. More precisely, these tools and techniques have been designed to efficiently compare the impact between two versions of a security policy applied to the same workflow, one before, the other after a policy modification. This work focuses on two specific types of security impact analysis. The first is quantitative in nature, providing a measure of success rate for a security constrained workflow which must be executed by employees who may be absent at runtime. This work considers quantifying workflow resiliency which indicates a workflow’s expected success rate assuming the availability of employees to be probabilistic. New aspects of quantitative resiliency are introduced in the form of workflow metrics, and risk management techniques to manage workflows that must work with a resiliency below acceptable levels. Defining these risk management techniques has led to exploring the reduction of resiliency computation time and analysing resiliency in workflows with choice. The second area of focus is more qualitative, in terms of facilitating analysis of how people are likely to behave in response to security and how that behaviour can impact the success rate of a workflow at a task level. Large amounts of information from disparate sources exists on human behavioural factors in a security setting which can be aligned with security standards and structured within a single ontology to form a knowledge base. Consultations with two CISOs have been conducted, whose responses have driven the implementation of two new tools, one graphical, the other Web-oriented allowing CISOs and human factors experts to record and incorporate their knowledge directly within an ontology. The ontology can be used by CISOs to assess the potential impact of changes made to a security policy and help devise behavioural controls to manage that impact. The two consulted CISOs have also carried out an evaluation of the Web-oriented tool. viii
Description: PhD Thesis
URI: http://hdl.handle.net/10443/3919
Appears in Collections:School of Computing Science

Files in This Item:
File Description SizeFormat 
Mace, J.C 2017.pdfThesis3.43 MBAdobe PDFView/Open
dspacelicence.pdfLicence43.82 kBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.