基于溯源图社群发现的APT行为检测
首发时间:2024-01-23
摘要:高级持续威胁(Advanced Persistent Threats,APT)是一种手段高超、潜伏时间长、危害性大的复杂攻击行为,因此及时发现其异常行为并进行阻断十分重要。本文对APT攻击不同阶段的节点的社群行为进行研究,基于在社交网络等领域有着出色表现的社群发现算法Louvain算法进行改进使其适用于APT攻击溯源图的社群发现,之后基于对APT攻击行为特点提取特征,并利用半监督KNN算法处理未标注数据,最后使用RIPPER算法实现节点APT异常行为检测。实验结果表明,本文提出的算法在溯源图社群划分效果与攻击行为检测准确度相较之前相关领域的研究有所提高。
关键词: 数据安全与计算机安全 APT攻击检测 社群发现
For information in English, please click here
APT attacks detection based on provenance graph community discovery
Abstract:Advanced Persistent Threats (APT) is a kind of complex attack behavior with sophisticated means, long latency and great harm. Therefore, it is very important to detect and block the abnormal behavior in time. In this paper, we studied the community behavior of nodes in different stages of APT attacks. Based on the community discovery algorithm with excellent performance in social networks and other fields, Louvain algorithm is improved to make it suitable for the community discovery of APT-attack provenance graph Then, based on the characteristics of APT attack behavior, features are extracted and unmarked data is processed by semi-supervised KNN algorithm. Finally, RIPPER algorithm is used to detect node APT abnormal behavior. The experimental results show that the algorithm proposed in this paper has improved the efficiency of community detection and the accuracy of attack behavior detection compared with previous studies in related fields.
Keywords: Data security and computer security Advanced Persistent Threats Detection Community Detection.
基金:
引用
No.****
同行评议
勘误表
基于溯源图社群发现的APT行为检测
评论
全部评论