A Network Security Situation Assessment Method Based on Fusion Model

This paper proposes a new model for network security situation assessment (NSSA) that addresses the limitations of current methods in terms of feature extraction quality and efficiency. The proposed NSSA model is based on a fusion model (FM) that incorporates an attention mechanism and bi-directional gated recurrent unit (BiGRU). The FM model is used to extract key information about different cyber threats, and the attention mechanism is applied to weight these key features, thereby improving the model's accuracy. Finally, the evaluation results are output by BiGRU, in combination with proposed quantitative indicators of cybersecurity posture. The results of threat detection experiments show that the proposed FM model outperforms other models based on several judging metrics.