摘要
云服务的文件存储存在“后门”攻击,以混淆用户视听,窃取用户隐私。现有的检测方法单一且需要更多的运行内存,因此文章提出通过使用AC自动机算法和朴素贝叶斯算法,快速精准地识别文本内容,利用scikit-learn机器学习库对图片内容进行甄别,且调用VirusTotal的API检测恶意文件,实验结果表明该检测方法在识别恶意内容的准确率上达到96.2%、可对海量数据进行实时检测,优于其他检测方法。
There are“backdoor”attacks on the file storage of cloud services to confuse visual and auditory sense of users and steal their privacy.The existing detection methods are single and require more running memory.Therefore,this paper proposes to use AC automaton algorithm and Naive Bayesian algorithm to identify text content quickly and accurately.It uses scikit-learn machine learning library to screen image content,and calls API of VirusTotal to detect malicious files.The experimental results show that the detection method achieves 96.2%accuracy in identifying malicious content and can detect massive data in real time,which is better than other detection methods.
作者
魏先燕
卢加奇
冯燕茹
吕广旭
王小英
WEI Xianyan;LU Jiaqi;FENG Yanru;LYU Guangxu;WANG Xiaoying(Institute of Disaster Prevention,Langfang 065201,China)
出处
《现代信息科技》
2023年第12期155-157,161,共4页
Modern Information Technology
基金
防灾科技学院教育研究与教学改革项目(JY2022B31)。
