有向切换网络有害信息智能过滤方法仿真

Simulation of Intelligent Filtering Method for Harmful Information in Directed Switching Networks

对有向切换网络的有害信息进行智能过滤时,在提高网络安全应用方面具有重要意义。在对网络有害信息进行过滤时,需对网络信息的特征权重进行计算,确定网络模式的隶属函数。传统方法主要根据网络信号进行过滤,忽略了隶属函数的影响,导致过滤有害信息的准确率低,提出基于模式影响度方法的有害信息过滤方法。根据网络信息分类的向量和特征权值,计算网络信息过滤的评判矩阵,得到网络智能信息向量和网络信息类别的相似度值,进一步对网络信息熵和条件信息熵进行求解,以网络有害信息的特征选择为基础,利用网络信息的特征项计算特征项在网络信息中的的频率权重和位置权重,分析有向切换网络模式的影响函数,分别计算网络信息模式库的隶属度函数,根据调整网络中的阈值和计算网络信息的正常度,实现有害信息的智能过滤。仿真结果表明,提出方法在对网络有害信息进行智能过滤时,具有较高的查全率和准确率。

This paper proposes a method of filtering harmful information based on the pattern influence degree.According to the vector and feature weight of network information classification,the judgment matrix of network information filtering was calculated to get the similarity values of network intelligent information vector and network information category. Furthermore,the network information entropy and the condition information entropy were solved.Based on the feature selection of harmful information of network,the feature term of network information was used to calculate the frequency weight and position weight of feature item in network information. In addition,the influence function of directed switching network mode was analyzed. Respectively,the membership function of network information pattern database was calculated. Finally,we achieved the intelligent filtering of harmful information by adjusting the threshold in network and calculating the normality degree of network information. Simulation results show that the proposed method has high recall rate and accuracy in intelligently filtering harmful network information.