Analýza bezpečnostních hrozeb s ohledem na ochranu soukromí

Indicators of Compromise, Federated Learning, Privacy preserving, Anomaly
detection, Cybersecurity

This project aims to research and develop a cybersecurity threat detection system
that emphasizes the collection of Indicators of Compromise (IoCs) from anomaly
detection systems placed in multiple customer networks, while ensuring the
privacy of the individuals who are monitored by those systems. The primary goal
is to increase the cybersecurity of end users through enhanced threat detection
capabilities of anomaly detection systems based on the cooperation of those
systems. The challenge, however, is to address the potential privacy concerns
associated with collecting and analyzing sensitive data such as IP addresses,
login credentials, and activity patterns. Such information can potentially be
misused and lead to privacy violations. Therefore, the project aims to protect
sensitive data and provide end users with a robust privacy guarantee to encourage
them to share detected security events enabling to determine IoCs for external
analysis and increasing threat detection.