Back
Analysis of malicious campaigns in multiple heterogeneous threat datasources
Michał Kruczkowski
Abstract
This doctoral dissertation concerns the problems of identification of malware (malicious software) campaigns on the Internet. This is an extremely important issue because it arises from a real need to ensure safety in rapidly growing computer networks. Proposed approach assumes the use of data mining and machine learning methods. It utilizes data about threat incidents t aken from multiple datasources related with v arious layers of ISO/OSI network communication model. The r esults of the doctoral dissertation confirm that the automated analysis and classification of data from heterogeneous datasources can be efficiently used to protect the Internet. The use of full info rmation about threats, including v arious network layers, leads to achieve better results compared with frequently used single layer analysis. Data mining and machine learning methods can efficiently support the cybercrime protection systems. The system MalCAS (Malware Campaign Analysis System) for malware campaigns identification that implements these methods fully corresponds the demands of contemporary networks. It can be applied as a useful and powerful tool to support the software environment ensuring net work cybersecurity- Record ID
- WUT16c6df045a4c4214bf32dcac181b35d1
- Diploma type
- Doctor of Philosophy
- Author
- Title in Polish
- Analiza złośliwych kampanii na podstawie danych o atakach pozyskiwanych z heterogenicznych zródeł
- Title in English
- Analysis of malicious campaigns in multiple heterogeneous threat datasources
- Language
- (pl) Polish
- Certifying University/Institution (when outside WUT)
- Systems Research Institute (IBS PAN) [Polish Academy of Sciences (PAN)]
- Discipline
- automation and robotics / (technology domain) / (technological sciences)
- Status
- Finished
- Defense Date
- 20-11-2015
- Title date
- 20-11-2015
- Supervisor
- External reviewers
- Keywords in English
- x
- Abstract in English
- This doctoral dissertation concerns the problems of identification of malware (malicious software) campaigns on the Internet. This is an extremely important issue because it arises from a real need to ensure safety in rapidly growing computer networks. Proposed approach assumes the use of data mining and machine learning methods. It utilizes data about threat incidents t aken from multiple datasources related with v arious layers of ISO/OSI network communication model. The r esults of the doctoral dissertation confirm that the automated analysis and classification of data from heterogeneous datasources can be efficiently used to protect the Internet. The use of full info rmation about threats, including v arious network layers, leads to achieve better results compared with frequently used single layer analysis. Data mining and machine learning methods can efficiently support the cybercrime protection systems. The system MalCAS (Malware Campaign Analysis System) for malware campaigns identification that implements these methods fully corresponds the demands of contemporary networks. It can be applied as a useful and powerful tool to support the software environment ensuring net work cybersecurity
- Uniform Resource Identifier
- https://repo.pw.edu.pl/info/phd/WUT16c6df045a4c4214bf32dcac181b35d1/
- URN
urn:pw-repo:WUT16c6df045a4c4214bf32dcac181b35d1