Risk Analysis of Enterprise Cross-Network Data via DAE-BiLSTM and Fuzzy Bayesian Network

With the increasing severity of enterprise information security issues, the identification and evaluation of cross-network data risk has become one of the essential technologies to safeguard enterprise data and network security. However, traditional rule-based or signature-based anomaly detection methods are increasingly unable to cope with the complex and evolving forms of attacks, necessitating more intelligent and flexible detection mechanisms to counter these constantly changing threats. To address this, a method integrating denoising autoencoder (DAE) and bidirectional long short-term memory network (BiLSTM) is first proposed for monitoring the enterprise cross-network data traffic and identify the occurring abnormal risks. On this basis, we use a fuzzy Bayesian network to effectively measure these known abnormal risks. In the end, experimental results demonstrate that the proposed method outperforms traditional machine learning algorithms and other deep learning models across various evaluation metrics, exhibiting high accuracy and efficiency.