基于知识图谱的恶意域名检测方法

人工智能在恶意域名检测领域的应用越来越广泛,而传统的恶意域名检测方法主要采用黑名单方式,存在时效性较差的问题。因此,提出了一种将知识图谱与恶意域名检测相结合的系统,完成了信息在知识图谱中的存储和表示。将系统的嵌入式模型作为输入,使用BiLSTM神经网络提取特征并完成最终的检测。实验表明,在通过真实数据构造的数据集上,该系统性能良好,对恶意域名的检测准确率高达99.31%。

Detection Method of Malicious Domain Name based on Knowledge Map

The application of artificial intelligence in the field of malicious domain name detection is becoming more and more widespread.However,the traditional malicious domain name detection method mainly uses the blacklist method,which has the problem of poor timeliness.Therefore,a system combining knowledge map with detection of malicious domain names is proposed to complete the storage and representation of information in the knowledge map.Taking the embedded model of the system as input,BiLSTM neural network is used to extract features and complete the final detection.Experiments indicate that the system performs well on datasets constructed from real data,and the accuracy of detecting malicious domain names is as high as 99.31%.