基于改进DS证据理论和BN的信息安全风险评估

针对信息安全风险评估过程中专家评价意见的多样性以及不确定信息难以量化处理的问题,提出了一种基于改进的DS证据理论与贝叶斯网络(BN)结合的风险评估方法.首先,在充分研究信息安全风险评估流程和要素的基础上,建立了风险评估模型,确定风险影响因素;其次,根据评估模型并结合专家知识构建相应的贝叶斯网络模型,确定贝叶斯网络模型中的条件概率表;再次,利用基于权值分配和矩阵分析的改进DS证据理论融合多位专家对风险影响因素的评价意见;最后,根据贝叶斯网络模型的推理算法,计算被测信息系统处于不同风险等级的概率值,并对结果进行有效性分析.分析表明,将改进后的DS证据理论与贝叶斯网络应用到风险评估过程中,在一定程度上能够提高评估结果的可信度和直观性.

Information security risk assessment based on the improved DS evidence theory and BN

For the diversity evaluations from experts and uncertain information is difficult to quantify,this paper proposes a risk assessment method on the basis of improved DS evidence theory and Bayesian network.Firstly,based on the processes and elements of the information security risk assessment,an assessment model is established and risk factors are confirmed.Secondly,according to the assessment model and expert knowledge,we establish Bayesian network model and determine the conditional probability table.Thirdly,the experts evaluations are integrated through a improved D-S evidence synthesis algorithm based on matrix analysis and weight distribution.Finally,combining with the inference procedure of Bayesian network,we calculate the probability of the information system is in different risk level and analysis results.The results show that it is more credible and intuitive when the proposed method which is based on improved DS evidence theory and Bayesian network is applied to risk assessment process.