Scholarlink

In response to the escalating threat posed by Advanced Persistent Threat (APT) attacks, there is an urgent demand for robust detection mechanisms. This paper introduces a pioneering approach to APT detection, leveraging the transformative power of hypergraphs and enabling the application of Behavior Clustering techniques. By harnessing hypergraphs, our method transcends the limitations of traditional analysis by capturing intricate relationships among network entities. Our framework, HyperBC, represents a significant advancement by integrating hypergraph theory into APT detection methodologies. Through the construction of a hypergraph representation of network data, HyperBC offers a comprehensive understanding of APT behaviors, leading to enhanced detection accuracy and resilience against sophisticated attack strategies. Extensive experimentation on the UNSW dataset and existing datasets validates the efficacy of our approach in identifying and mitigating APT threats, thus making a substantial contribution to cybersecurity research and fortifying defenses against evolving cyber threats.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

HyperBC: Hypergraph-Based Approach for Behavior Cluster of Suspicious APT Attacks

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Wenhui Du,
  • Shuilin Li,
  • Gaolei Li,
  • Jianhua Li

摘要

In response to the escalating threat posed by Advanced Persistent Threat (APT) attacks, there is an urgent demand for robust detection mechanisms. This paper introduces a pioneering approach to APT detection, leveraging the transformative power of hypergraphs and enabling the application of Behavior Clustering techniques. By harnessing hypergraphs, our method transcends the limitations of traditional analysis by capturing intricate relationships among network entities. Our framework, HyperBC, represents a significant advancement by integrating hypergraph theory into APT detection methodologies. Through the construction of a hypergraph representation of network data, HyperBC offers a comprehensive understanding of APT behaviors, leading to enhanced detection accuracy and resilience against sophisticated attack strategies. Extensive experimentation on the UNSW dataset and existing datasets validates the efficacy of our approach in identifying and mitigating APT threats, thus making a substantial contribution to cybersecurity research and fortifying defenses against evolving cyber threats.