【作者】 江华丽；

【导师】 王平；

【作者基本信息】 福建师范大学 ， 通信与信息系统， 2010， 硕士

【摘要】 目前关于垃圾邮件过滤的研究技术同垃圾邮件制造者的技术存在着激烈的竞争,邮件给人们的日常生活和工作带来很大便利,但是同时垃圾邮件的涌现和病毒的侵扰也带来了一系列的负面影响,因此垃圾邮件的泛滥需要迫切的技术处理。本次设计基于前人研究的基础上,结合邮件知识和神经网络研究了一套邮件过滤系统,采用人工智能的方法提取邮件特征,自动学习垃圾邮件的特征并进行过滤,提高邮件过滤的准确度,最后并将邮件过滤系统应用到嵌入式系统,这样的嵌入式设备可以随身携带,可以方便的进行通信。本文首先阐述了邮件的工作原理及基础知识,再通过对神经网络算法的研究并算法进行改进,将邮件过滤与神经网络相结合,提取垃圾邮件特征并进行训练和测试,最后阐述移植的过程,将完整的系统移植到嵌入式Windows CE操作系统平台上,通过实验验证设计的可行性。更多还原

【Abstract】 At present, the spam filtering technology and the study of technology spammers are constantly competing with each other, e-mails facilitate people’s daily lives and the work, but the emergence of spam and the virus has also brought a series of intrusive negative impact, so the popularity of spam technology requires urgent treatment. The design is based on the basis of previous studies, combining with e-mail knowledge and neural network to design a set of e-mail filtering system, and use the artificial intelligence to extract the e-mail features, this system can learn the spam’s characteristics automatically and improve the accuracy of mail filtering. Finally, the e-mail filtering system will be ported to the embed system, which is portable and can communicate easily. This article tells the working principle and basic knowledge of email firstly, then combines the neural network algorithm with the e-mail filtering system, after extracting the characteristics of junk e-mail, and then train and test it.Finally, explaining the transplant process, the whole system will be completely transferred to the embedded Windows CE operating system platform, experiments can demonstrate the feasibility of the design.更多还原