Back to search

Democritus University of Thrace (DUTH)

Dynamic risk assessment in cybersecurity

Abstract

dc:description

Traditional information security risk assessment methodologies and standards, which serve as foundational components of information security management systems, face significant challenges in contemporary environments. These challenges stem from the rapidly evolving threat landscape and the continuous emergence of new vulnerabilities. To address these issues, dynamic risk assessment models have been introduced, enabling continuous and near real-time evaluation of risks to organisational operations. This doctoral thesis critically examines the current state of dynamic risk assessment models in cybersecurity and proposes novel approaches to advance existing methodologies. The research begins with a systematic literature review that explores dynamic risk assessment models within the cybersecurity domain. Through a rigorous screening process, fifty models were identified, thoroughly examined, and analysed. The review uncovers findings related to their primary analysis methods and the domains or application areas in which these models flourish. Additionally, it provides insights into key characteristics, including maturity levels and the types of information these models utilize to generate results. A significant finding is that most models operate as comprehensive estimation frameworks, primarily emphasizing changes in the probability or likelihood of threat occurrences. These frameworks are based on the fundamental assumption that risk is a function of both probability and impact. However, a notable limitation is the lack of proactive capabilities due to limited integration of relevant data. Most existing models primarily capture real-time information reflecting ongoing activities within an organization’s environment, which restricts their capacity to anticipate and preemptively address emerging threats. To address these limitations and advance the field of dynamic risk assessment in cybersecurity, this thesis introduces four novel methodologies. The first is the Dynamic Vulnerability Severity Calculator, which dynamically updates the vulnerability vector within the risk assessment formula. This approach incorporates analyses of environmental topology and the effectiveness of deployed security mechanisms. It further leverages the Common Vulnerability Scoring System framework to adjust the severity of detected vulnerabilities based on the specific organisational environment. In addition to evaluating individual vulnerabilities, the methodology considers their quantity and interdependencies within each asset for a more precise assessment. These factors are integrated into a comprehensive Fuzzy Cognitive Map model, which includes attack paths to derive an overall vulnerability score. A detailed case study in a SCADA environment, supported by modified scenarios, validates the practical applicability and effectiveness of this approach. The second proposed methodology presents a proactive and dynamic risk assessment model that combines data from an organisation’s internal environment with relevant cybersecurity-related open sources. Qualitative data are converted into numerical form and integrated with quantitative data using a structured methodology. This combined information is incorporated into a Bayesian Network model, which dynamically estimates the probability of a cyber attack’s success. When coupled with impact assessments of organisational assets, the model delivers comprehensive risk estimations. By integrating the Exploit Prediction Scoring System, this methodology offers risk assessments that are not only dynamic but also proactive. Furthermore, the model considers the influence of Advanced Persistent Threat on risk estimation. The effectiveness of this model is validated through a detailed case study, demonstrating its application in assessing cyber risks within a Supervisory Control and Data Acquisition environment. The third methodology presents an integrated approach that unifies the Dynamic Vulnerability Severity Calculator and the Proactive and Dynamic Threat Assessment into a cohesive attack scenario framework. This model retrieves cybersecurity-related data through vulnerability detection and utilizes it for dynamic risk estimation. Upon successful risk assessment, it provides information regarding available patches for detected vulnerabilities and offers mitigation recommendations to proactively reduce risks. The effectiveness of this integrated model is validated through a case study within a small network environment. The final model advances risk assessment by introducing time-sensitive analysis alongside proactive and dynamic capabilities. This methodology leverages the Exploit Prediction Scoring System to estimate the short-term likelihood of exploitation over a 30-day period. To enhance accuracy, Bayesian networks are employed to model system vulnerabilities and asset interdependencies within the network. This information is integrated into an absorbing Markov chain, incorporating identified attack paths explored through Depth-First Search. The model generates exploitation probability distributions over a predefined time window, which, when combined with asset impact assessments, facilitates dynamic, proactive, and time-sensitive risk estimations. Additionally, it provides critical insights into attack progression by estimating the time required for an adversary to compromise key assets. A case study demonstrates the model’s practical application in assessing cyber risks within a SCADA environment. Overall, this thesis aspires to contribute to the scientific community by providing a systematic literature review and introducing novel dynamic risk assessment models. These contributions address the limitations of existing approaches and offer enhanced methodologies to strengthen cybersecurity posture of organisations.

Degree

thesis:*
Grantor dc:publisher
Democritus University of Thrace (DUTH)
Year dc:date
2025

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Cheimonidis, Pavlos
  • Χειμωνίδης, Παύλος

Subjects

dc:subject × 20

Rights

Language dc:language
eng

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/59753

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Cheimonidis, Pavlos; Χειμωνίδης, Παύλος. Dynamic risk assessment in cybersecurity. Democritus University of Thrace (DUTH), 2025. http://hdl.handle.net/10442/hedi/59753