Vulnerability assessment is a foundational and important work for network security.There are many existing vulnerability evaluation algorithms,but most of them are based on fixed formu-las,so it is often unable to conduct real-time and dynamic evaluation on vulnerabilities.By analyzing the existing problems and deficiencies of the current vulnerability evaluation methods,this paper propo-ses the basic principle and algorithm flow based on bidirectional LSTM and SVM,including data pre-processing,feature selection,model construction,model construction and experimental verification.The experimental results show that BiLSTM-SVM can accurately classify and predict vulnerabilities,improve the accuracy and efficiency of vulnerability evaluation,and thus provide a powerful and more ef-fective assessment method support for vulnerability management.
万方数据
维普
