面向网络状态的自适应用户行为评估方法

用户复杂的行为往往会导致网络状态出现波动，破坏网络平稳运行，由于指标及权重的主观性和静态性，传统评估方法难以准确衡量用户行为对网络状态变化的影响程度。因此引入粗糙集理论，构建一种面向网络状态的自适应用户行为评估方法，使用属性约简和属性重要度方法对用户行为和网络状态数据进行挖掘，分析用户行为与网络状态变化的关联程度，以此自适应构建评估指标及权重，并随用户行为变化而动态调整，从而准确地量化用户行为对网络状态变化的影响程度。实验结果表明，该评估方法有助于准确发现造成网络状态变化的用户及其行为，能够为加强对用户行为的管控提供有效支持。

Adaptive user behavior’s evaluation method based on network status

Abstract:With the development of network, user's complex and dynamic behaviors often lead to unexpected fluctuations of network status,and these fluctuations take great challenge to the stability of network. It is difficult to evaluate user be-havior's affection on network status fluctuations for traditional methods because of their subjectivity and static drawbacks. An evaluation method was proposed, which analyses the correlation between the user's behavior and network status fluc-tuations based on actual data of behavior and network status,by means of rough set attribute reduction and attribute im-portance. This method can construct and dynamically adjust the evaluation indexes and their weight adaptive the actual data. The evaluation results show this method can help to detect and manage users who affect the stability of net-work, and provide effective support to control users and their beha ior.