Graph Convolutional Networks for Malicious URLs Detection
Huanyu Yang, Gengsheng Zheng · 2024
This study aims to explore the application of graph convolutional networks in the task of malicious domain detection, integrating it with cybersecurity issues. By leveraging graph convolutional networks to model the relationships between domains, we can more accurately capture semantic information and structural features among domains, thereby improving detection accuracy and generalization ability. The model achieves an average prediction accuracy of over 91.27% on two different datasets, malicious-url and CSIC. Experimental results demonstrate significant performance improvement in the task of malicious domain detection, validating the effectiveness of combining graph convolutional networks with cybersecurity problems.