Security Alert Priority Determination via Traffic Dissimilarity Index

The patent applies dynamics by transitioning from static pre-defined events to dynamic, adaptive event definitions. The system learns from historical traffic flow data and automatically updates event patterns, allowing the priority determination mechanism to adapt to new attack types and evolve over time without manual reconfiguration, thus maintaining both simplicity and accuracy

The patent changes parameters by using multiple attributes of traffic flow (packet size, frequency, timing patterns, protocol types) instead of relying on a single pre-defined event category. This multi-parameter approach enables more nuanced priority determination that captures the complexity of modern cyber threats while maintaining an automated process

The patent applies preliminary action by pre-processing and storing historical traffic flow data in a structured manner before it is needed for priority determination. The system performs preliminary learning and pattern recognition on historical data, building a knowledge base that enables accurate and adaptive priority assessment when new security alerts occur

The patent implements feedback mechanisms where the priority determination results and analyzed traffic patterns are fed back into the system to continuously refine event definitions and improve future determinations. This closed-loop approach enables the system to adapt to new attack types while maintaining and improving accuracy over time through iterative learning