Detection of Botnet Malware using AI-based Keystroke Dynamics (DEAL 665)

This lot relates to a technique for detecting stealthy malware that hides its activity within normal-looking network traffic, using machine learning algorithms to protect the network and achieve high accuracy by monitoring multiple network events and identifying the legitimacy of root-trigger events, thereby reducing false positives in the network. The system converts individual network events and root-trigger events into event pairs with comparable pairwise attributes, such as similarities in event timestamps, domain names, host names, IP addresses, process IDs (PIDs), HTTP request referrer fields, network request types, network request content, and network request flags. This technology may be implemented in network traffic management systems, enterprise network security systems, military cyber defense systems, stealthy attack detection systems, etc.