摘要
针对传统校园网络安全态势预测误差大的情况,设计了一种基于无监督聚类算法的校园网络安全态势预测方法。应用无监督聚类算法,确定校园网络中的隐藏节点,根据无监督聚类结果,提取安全态势要素,将要素规范处理后得到安全态势要素集。在此基础上,使用安全态势要素集从时间上预测未来时间内的安全态势要素,判断未来时间的网络安全态势情况,由于时间预测具有一定的不确定性,引入空间数据发掘理论,得到空间特性引发的安全态势要素之间的影响关系,并采用空间数据中的指标定量计算网络节点的距离关系,得到网络安全态势预测值,以此完成校园网络安全态势的预测。实验对比结果表明,此次设计的基于无监督聚类算法的校园网络安全态势预测方法比传统方法预测误差小,能够真实反应校园网络的安全态势情况。
In view of the large error of traditional campus network security situation prediction,a campus network security situation prediction method based on unsupervised clustering algorithm is designed.Applying unsupervised clustering algorithm,the hidden nodes in campus network are determined.According to unsupervised clustering results,security situation elements are extracted,and security situation elements set is obtained by normalizing the elements.On this basis,the security situation factor set is used to predict the security situation factor in the future time and judge the network security situation in the future time.Because the time prediction is uncertain,the spatial data mining theory is introduced to obtain the influence relationship between the security situation factor caused by the spatial characteristics.And the distance relationship of network nodes is quantitatively calculated by using spatial data indicators,and the network security situation prediction value is obtained,so as to complete the campus network security situation prediction.The experimental results show that the proposed campus network security situation prediction method based on unsupervised clustering algorithm has smaller prediction error than the traditional method,and can truly reflect the security situation of campus network.
作者
方晖
FANG Hui(Educational Technology Center of Huizhou University,Guangdong Huizhou 516007,China)
出处
《新一代信息技术》
2019年第19期12-16,29,共6页
New Generation of Information Technology
基金
广东省教育厅“基于‘核心竞争力’理论的应用型大学人才培养研究—以广东省经管类专业为例”(项目编号:2017ZX021)。
关键词
无监督聚类算法
网络安全
态势预测
空间
unsupervised clustering algorithm
network security
situation prediction
space
作者简介
方晖(1976-),男,硕士,计算机应用高级实验师,研究方向:软件技术、教育技术,教学资源课程建设,信息网络安全。
