基于云模型与贝叶斯反馈的网络安全等级评估方法
Evaluation method for network security level based on cloud model and Bayesian feedback

摘要： 鉴于测评人员对信息系统等级保护测评中网络安全评估的测评指标理解的局限性、评判的模糊性以及量化打分中判定区间的不确定性,将云模型与贝叶斯反馈算法相结合,采用云模型理论来处理评估结果的模糊性和随机性。由于传统的一维逆向云算法出现超熵为虚数的情况,因此,使用无确定度逆向云算法来避免这种情况的发生。通过建立贝叶斯反馈云模型,对给出的测评结果进行检验和修正,使评估结果更加客观,准确。根据已修正网络安全整体评估云的期望和熵,给出合理的安全等级判定区间,作为实际测评中等级评估的依据。实例验证表明,该评估方法可以有效地修正测评人员评估结果之间的差异,实现合理的判定划分,给出客观、准确的等级评估结果。
Abstract: In view of the limitations of the evaluation staff on the content of the evaluation index and the fuzziness of judgment and the limitation of the judgment interval in quantitative evaluation of network security evaluation in information system level protection. The researchers combine the cloud model with the Bayesian feedback algorithm and adopt the cloud model theory to deal with the fuzziness and randomness of the evaluation results. Because the traditional one-dimensional reverse cloud algorithm appears hyper entropy is imaginary, therefore, researchers use the uncertainty of the reverse cloud algorithm to avoid the occurrence of negative entropy. Through the construction of Bayesian feedback cloud model, which is constructed to test and correct the evaluation results given by experts or reviewers which makes the final evaluation results more objective and accurate. According to the expectation and entropy of the overall correction cloud of network security and the reasonable security grade judgment interval is given as the basis of the grade evaluation in the actual evaluation. Verified by an example, the evaluation method proposed that can effectively correct the difference between the evaluation results of the evaluation personnel, realize the reasonable judgment division and give the objective and accurate grade evaluation result