arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2601.04243cs.CRcs.AI
收藏

融合多智能体仿真、行为取证和基于信任的机器学习的自适应内部威胁检测

A Hybrid Insider Threat Detection Framework Combining Multi-Agent Simulation, Layered SIEM Correlation, and Theory-of-Mind Reasoning

Department of Computer Science(计算机科学系)

Firdous Kausar, Asmah Muallem, Naw Safrin Sattar, Mohamed Zakaria Kurdi

AI总结:

本文提出融合多智能体仿真、行为取证和基于信任的机器学习的自适应内部威胁检测框架,通过不同系统变体验证了认知上下文、证据门控验证和预训练通信校准对提升检测精度和降低误报的作用。

摘要:

我们提出了一种混合框架,用于自适应内部威胁检测,该框架紧密整合了多智能体仿真(MAS)、分层安全信息和事件管理(SIEM)相关性、行为和通信取证、基于信任的机器学习以及理论ofmind(ToM)推理。智能代理在模拟的企业环境中运行,生成行为事件和认知意图信号,这些信号被中央SIEM接收。我们评估了四个系统变体:分层SIEM核心(LSC)基线、结合ToM和通信取证的认知增强SIEM(CE-SIEM)、引入精确聚焦验证机制的证据门控SIEM(EG-SIEM),以及通过预训练邮件取证模块校准于Enron语料库的Enron增强EG-SIEM(EG-SIEM-Enron)。在涉及八名恶意内部人员的十次模拟运行中,CE-SIEM实现了完美召回率(1.000),并将Actor级F1值从0.521(LSC)提高到0.774。EG-SIEM将Actor级F1值提升到0.922,确认警报的精确度提高到0.997,同时将误报减少到每运行0.2个。EG-SIEM-Enron保持高精确度(1.000确认警报精确度;每运行0个误报),略微提高Actor级F1值至0.933,并减少检测延迟(平均TTD 10.26步,相较于EG-SIEM的15.20步)。这些结果表明,认知上下文提高了灵敏度,证据门控验证能够实现高精度、低噪声检测,预训练的通信校准可以进一步加速高置信度的内部威胁识别。

英文摘要:

This paper presents a hybrid insider threat detection framework for enterprise environments, integrating multi-agent simulation, layered SIEM correlation, trust-adaptive thresholds, behavioral and communication forensics, and Theory-of-Mind reasoning. Email is treated not as a control channel but as a coordination and social-engineering evidence stream correlated with authentication, file-access, and privilege events. Four variants are evaluated: Layered SIEM-Core (LSC), Cognitive-Enriched SIEM (CE-SIEM), Evidence-Gated SIEM (EG-SIEM), and EG-SIEM-Enron with Enron-calibrated email forensics. Across ten matched runs with eight malicious insiders, actor-level F1 improves from 0.567 for LSC to 0.774 for CE-SIEM, 0.898 for EG-SIEM, and 0.944 for EG-SIEM-Enron; paired Wilcoxon tests confirm the first three differences after Holm-Bonferroni correction. Evidence gating reduces confirmed false positives from 33.7 per run under LSC and 49.3 under CE-SIEM to 0.2 and 0.0 respectively, a precision gain traded against longer confirmation time. Domain-shift evaluation shows that an Enron-trained email classifier does not transfer to a different operational email domain, although target-domain fine-tuning reaches F1 = 0.974 under grouped template-family evaluation. On CERT r4.2, the evidence-gated logic improves actor-level detection while reducing false positives, outperforming tabular anomaly detectors and a sentence-embedding baseline. Scalability tests to 1000 agents indicate stable detection quality.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑