Přístupnostní navigace
E-application
Search Search Close
Doctoral Thesis
Author of thesis: Ing. Yehor Safonov, Ph.D.
Acad. year: 2026/2027
Supervisor: doc. Ing. Zdeněk Martinásek, Ph.D.
Reviewers: doc. Ing. Petr Františ, Ph.D., doc. Ing. Vladimír Soběslav, Ph.D.
In modern cybersecurity environments, Security Information and Event Management (SIEM) platforms form a central layer for security monitoring and Security Operations Center (SOC) operations. They collect, structure, analyze, and correlate security-relevant telemetry from operating systems, network devices, cloud services, security tools, applications, and external sources. Beyond log collection and alerting, mature SIEM environments also accumulate operational security knowledge, including detection logic, field mappings, correlation rules, investigation procedures, tuning decisions, and platform-specific configurations. However, the automation of security monitoring remains constrained by the heterogeneous and inconsistent nature of security logs. Incomplete parsing, weak semantic context, vendor-specific schemas, and fragmented operational knowledge reduce the reliability of detection, correlation, and incident investigation. These limitations increase the amount of repetitive manual work required from SOC analysts and slow down the effective use of collected telemetry. This doctoral thesis focuses on Artificial Intelligence Methods for Security Monitoring Automation. It investigates how machine learning, natural language processing, and AI-based log understanding can transform raw and semi-structured security logs into normalized, semantically enriched, and machine-processable representations suitable for SIEM and SOC workflows. The research addresses the data-understanding layer of security monitoring and analyzes SIEM and SOC environments as an interconnected ecosystem of telemetry sources, processing pipelines, operational procedures, and security knowledge objects. The proposed approaches combine domain-specific preprocessing, standardized log metakeys, adaptive learning mechanisms, human-in-the-loop validation, and AI-assisted semantic interpretation of security events. Selected parts of the thesis also examine the portability of operational security knowledge across SIEM environments, including detection logic, correlation rules, field mappings, and investigation patterns. Particular emphasis is placed on trustworthy AI principles, especially transparency, explainability, human oversight, auditability, robustness, accuracy, and operational reliability. The main contribution of the thesis is a conceptual and technical framework for applying artificial intelligence methods to security monitoring automation. By improving data quality, reducing manual effort, supporting SOC workflow automation, and enabling partial preservation and transfer of operational security knowledge, the thesis contributes to more scalable, interoperable, and trustworthy security monitoring in modern cybersecurity environments.
security monitoring automation, log parsing, log-data understanding, SIEM, SOC, machine learning, artificial intelligence
Date of defence
23.10.2026
Result of the defence
Defended (thesis was successfully defended)
Process of defence
Předseda komise uvítal členy a hosty obhajoby a vyzval Ing. Safonova k prezentaci výsledků své disertační práce. Během prezentace Ing. Safonov uvedl motivaci práce a její cíle. Pozornost pak soustředil na konkrétní přínosy své práce. Po vyjádření školitele a přečtení posudků oponentů práce konstatující splnění cílů disertační práce, následovalo nejprve vypořádání otázek oponentů. Do diskuze se pak zapojili i ostatní členové komise a přítomní hosté. Na položené otázky Ing. Safonov vždy reagoval pohotově a z diskuze vyplynuly i nové možné směry pokračujícího výzkumu v dané problematice. Celková doba obhajoby včetně diskuze trvala více než 1 hodinu. Následovalo neveřejné jednání členů komise, která tajně hlasovala.
Language of thesis
English
Faculty
Fakulta elektrotechniky a komunikačních technologií
Department
Department of Telecommunications
Study programme
Information Security (DPC-IBE)
Composition of Committee
prof. Ing. Jaroslav Koton, Ph.D. (předseda) pplk. Ing. Aneta Coufalíková, Ph.D. (člen) Ing. Patrik Morávek, Ph.D. (člen) Ing. Tomáš Mácha, Ph.D. (člen) Ing. Ondřej Krajsa, Ph.D. (člen) doc. Ing. Petr Mlýnek, Ph.D. (člen) doc. Mgr. Josef Horálek, Ph.D. (člen) Ing. Tereza Otčenášková, Ph.D., BA (člen) doc. Ing. Petr Františ, Ph.D. (člen) doc. Ing. Vladimír Soběslav, Ph.D. (člen)
Supervisor’s reportdoc. Ing. Zdeněk Martinásek, Ph.D.
Reviewer’s reportdoc. Ing. Petr Františ, Ph.D.
Reviewer’s reportdoc. Ing. Vladimír Soběslav, Ph.D.
Responsibility: Mgr. et Mgr. Hana Odstrčilová