网络伪装不良信息检测方法的研究与仿真

研究网络中不良文字信息检测问题,提高检测的准确率。针对传统的不良信息检测方法都是针对具体的非法汉字进行对比检测的,没有考虑到汉字中的语义特征,当不良信息由合法汉字组成的时候,基于特征的检测方法由于没有考虑语义的因素,过于依赖不良汉字库,造成不良信息漏检率很高的问题。为解决上述问题,提出一种根据语义关联决策的信息过滤技术,通过计算信息语义与不良信息语义的关联程度,运用语义因素判定非法信息,有效克服传统方法的弊端。实验证明,方法能快速、完整地将高度伪装的不良信息检测出来,保证了信息的安全,取得了不错的效果。

Network Camouflage Bad Information Detection Method of Research and Simulation

Research information detection of bad words in networks and improve the testing accuracy.Traditional bad information detection methods contrast the specific illegal Chinese characters with illegal information Chinese character library,and does not considered the semantic features of Chinese characters.When bad information expressed in legal Chinese characters,the miss detection rate is very high.Thfe paper proposed an information filtering technology based on the semantic association decision.Through calculation the association degree of information semantic and bad information semantic,the semantic factors were used to determine illegal information,which overcomes the disadvantages of traditional methods.The experimental results prove that the method can quickly find out disguised bad information,and ensure the security of information.