关键词: 通用漏洞评估系统（CVSS）, 指标量化, 反向传播（BP）神经网络, 评估模型, MATLAB

Abstract: Considering that there are several drawbacks included in CVSS 2.0, such as strongly subjectivity, inefficient maneuverability, the difficulty to create automated assessment model, the evaluation index system is improved based on CVSS 2.0 evaluation system. And the evaluation index system is divided into two parts which are objective category and subjective category. It optimizes evaluation factor with principles of BP neural network self-learning and builds an automation evaluation model based on BP neural network, then quantizes the input indicators characteristic into approximation of effectiveness rapidly. Finally the effectiveness, accuracy and feasibility of the method are proved by MATLAB simulation.

Key words: Common Vulnerability Scoring System（CVSS）, indicator quantified, Back Propagation（BP） neural network, evaluation model, MATLAB