中文摘要：

针对内网安全评估问题构建了内网安全风险评估的系统模型。该模型采用的是基于模糊综合评价的评估方法，并在此基础上重点讨论了基于聚类分析的风险分类方法，其摒弃了传统标准分类的主观方法，对大量客观实验数据进行聚类分析，使风险的分类更具客观性。同时，在评估过程中利用信息熵原理确定各风险因素的权重系数，避免直接赋值的主观性，从而使风险评估的整个过程更加客观，评估结果更加准确。

英文摘要：

An security risk assessment model in Intranet is established, because of the network assessment problems. It is implemented with the assessment method, which is based on fuzzy comprehensive evaluation. And on this basis, the risk classification based on cluster analysis method which abandoned the traditional methods is mainly researched and the risk based on clustering analysis method is discussed, so that risk assessment is more objective. Meanwhile, in the evaluation process using the information entropy theory to de- termine the weight coefficient of risk factors, avoid the subjectivity of direct assignment, thus make the whole process of assessment of the risks more objective and results more accurate.