垃圾邮件优化过滤方法的研究与仿真

发布时间：2019-03-25 13:47

【摘要】：研究网络通信垃圾邮件过滤优化问题,由于存在随机性,传统的垃圾邮件过滤方法均以内容过滤为主,未考虑其它因素,邮件内容过多会造成确认特征过多,而内容过少则会影响确认的准确性,未全面考虑其它可识别特在因素,具有检测效率低和误差率高的缺陷。提出一种基于Kalmann滤波器和主元特征相似判断模型的垃圾邮件过滤方法,提取邮件的结构主元特征,采用Kalmann滤波器的权值更新模型调整特征项的权重,如果出现新信息,则运算全部特征相权重和,并将其转化成概率值,若概率值高于设置的阀值,通过主元相似判断模型分析待识别邮件,准确过滤相应的垃圾邮件。实验结果表明,改进方法能够具有较高的垃圾邮件识别效率和精确度,具有较强的应用价值。
[Abstract]:The optimization of spam filtering in network communication is studied. Due to the randomness, the traditional spam filtering methods are mainly based on content filtering, without considering other factors, too much email content will lead to too many confirmation characteristics. However, too little content will affect the accuracy of the confirmation, not taking into account other identifiable specific factors, which has the defects of low detection efficiency and high error rate. This paper proposes a spam filtering method based on Kalmann filter and principal component feature similarity judgment model, extracts the structural principal component feature of e-mail, adjusts the weight of feature item by the weight update model of Kalmann filter, if new information appears, Then the sum of all characteristic phase weights is calculated and converted into probability value. If the probability value is higher than the set threshold, the message to be identified is analyzed by principal component similarity judgment model, and the corresponding spam e-mail is filtered accurately. The experimental results show that the improved method has high efficiency and accuracy of spam recognition and has strong application value.
【作者单位】：南通大学现代教育技术中心;
【基金】：国家自然科学基金项目(61171132) 江苏省自然科学基金项目(BK2010280) 南通市应用研究计划项目课题(BK2011003,BK2012034)
【分类号】：TP393.098

【相似文献】