Identify Chinese Unsolicited Commercial Email (UCE) with Artificial Neural Networks

摘要：

With Email used widely,the rate of Internet is blocked by the flood of Unsolicited Commercial Email (UCE). It will maybe waste some time to deal with the UCE for users. Thus,its necessary to design a personal filter to identify UCE. So,a procedure of preprocess is briefly introduced in the paper. Then an Email filter is designed with BP neural networks. The filter identified UCE by its content. According to every user,this system learned his personal settings by itself. Consequently,it succeeded in realizing the automatic filter of the Chinese UCE. This system was tested with actual UCE and simulated in MATLAB6.5. The highest Spam precision is 95％ and average Spam recall is 68％.

关键词： Email filter BPNN CHINESE UCE UCE filter Vector Space Model Introduction

作者: Sun Yuhai Lu Jianhua Li Yi Lu Sida

作者单位: Artillery Commanding Academy,Langfang China Northwest A&F University,Yangling China

会议类型: 国际会议

会议名称: 2011 International Conference on Opto-Electronics Engineering and Information Science(2011光电电子工程与信息科学国际会议 ICOEIS 2011)

会议地点: 西安

会议语种:英文

页码: 2504-2508

在线出版日期: 2011-12-23（万方平台首次上网日期，不代表论文的发表时间）