一种基于行为链的恶意行为检测方法
A Malicious Behavior Detection Method Based on Behavior Chain
康铭伟 1宁达2
作者信息
- 1. 武汉市江夏区藏龙大道3号 武汉 430205
- 2. 武汉船舶通信研究所 武汉 430205
- 折叠
摘要
提出了一种基于行为链的恶意行为检测方法,通过API监控技术对行为点进行抽象,利用运行中的行为点调用顺序构建行为链,通过深度学习模型LSTM预测行为是否为恶意行为.实验结果表明,基于上述方法提取的行为点及构建的行为链在实验中有着很高的识别率,对于恶意行为检测具有较好的识别能力.
Abstract
A malicious behavior detection method based on behavior chain is proposed,which abstracts behavior points through API monitoring technology,constructs behavior chain using the calling order of running behavior points,and predicts whether the behavior is malicious through deep learning model LSTM.The experimental results show that the behavior points extract-ed and the constructed behavior chain based on the above methods have a high recognition rate in the experiment,and have good recognition ability for malicious behavior detection.
关键词
恶意行为/行为链/LSTMKey words
malicious behavior/behavior chain/LSTM引用本文复制引用
基金项目
海军装备"十四五"预研项目(3020909)
出版年
2024
国家科技期刊平台
NSTL
万方数据