一种基于最大频繁子图基因的模糊图神经网络检测模型  

Fuzzy graph neural network detection model based on maximum frequent subgraph genes

作　　者：周显春[1] 肖衡[1] 焦萍萍 邹琴琴 Zhou Xianchun;Xiao Heng;Jiao Pingping;Zou Qinqin(School of Information and Intelligence Engineering,Sanya University,Sanya,Hainan 572022,China)

机构地区：[1]三亚学院信息与智能工程学院,海南三亚572022

出　　处：《计算机时代》2023年第9期14-18,共5页Computer Era

基　　金：海南省自然科学基金项目资助(620MS064);三亚市院地科技合作项目资助(2019YD26);三亚学院优势专业建设项目(SYJZUS202203);三亚学院一流本科专业特色建设资助项目(SYJZZZ202212)。

摘　　要：针对恶意软件检测的准确性和时间效率问题,提出一种基于最大频繁子图基因的模糊图神经网络检测模型。首先利用SFFSM-SPIN-MGM方法挖掘恶意软件函数调用图的最大频繁子图,然后利用模糊图神经网络完成恶意软件同源性检测。实验结果表明,该方法具有较强的泛化能力,能够有效地检测现有恶意软件的变种测试集,平均准确率92.1%,平均误报率4.3%、平均漏报率1.4%。Aiming at the accuracy and time efficiency of malware detection,a fuzzy graph neural network detection model based on the maximum frequent subgraph genes is proposed.It first utilizes the SFFSM-SPIN-MGM method to mine the maximum frequent subgraphs of malware function call graphs,and then uses the fuzzy graph neural network to complete malware homology detection.Experimental results show that this method has strong generalization ability and can effectively detect existing malware variant test sets.The average accuracy rate of this model reaches 92.1%,the average false alarm rate is 4.3%,and the average miss rate is 1.4%.

关 键 词：恶意软件 动态函数调用图 最大频繁子图基因 模糊图神经网络 

分 类 号：TP309.5[自动化与计算机技术—计算机系统结构]