Abstract: To study the problem of network threat assessment, a new method based on attribute recognition theory was proposed. This method extracted indicators of network attacking events in a period and modeled with entropy value to determine weights. Threat degree was quantified after the assessment and classification of attacking events in the network. The example's results indicate that this method is simple, practical and able to assess network threats effectively.

Key words: network threat assessment, attribute recognition, entropy value, modeling

摘要： 针对网络威胁评估问题，提出一种基于属性识别理论的网络威胁评估方法。通过提取一段时间内网络攻击事件的指标，采用熵值确定权重并建立模型，对攻击事件进行评估、分级，进而量化出该段时间内网络的威胁程度。实例分析结果表明，该方法简单、实用，能够有效进行网络威胁的评估。

关键词: 网络威胁评估, 属性识别, 熵值, 建模