(电子科技大学 计算机科学与工程学院， 成都 610054)
摘 要：现有垃圾短信过滤系统主要采用对短信进行逐条分析判断的技术，因此处理的效率比较低。针对这一过滤技术的不足，提出了一个基于抽样的垃圾短信过滤方法，该方法引入用户信任度的概念，根据用户的信任度对用户发送的短信进行抽样过滤，极大地提高了处理效率。同时该方法整合了多项垃圾短信过滤技术（黑白名单、关键词及内容过滤技术），较之单一的过滤方法在准确率和效率上有很大的提高。
关键词：垃圾短信； 用户信任度； 抽样过滤； 文本分类
中图分类号：TP393 文献标志码：A
文章编号：10013695(2009)03093303
Filtering algorithm of junk SMS based on sample
ZHONG Yanhui, FU Yan, CHEN Anlong, GUAN Na
(School of Computer Science & Engineering, University of Electronic Science & Technology of China, Chengdu 610054, China)
Abstract:The existing filter system of junk SMS use the technology which judge SMS one by one, therefore its efficiency is quite low. To overcome the shortcomings of existing filtering technologies, this paper proposed a filtering algorithm of junk SMS based on sample. Introduced the concept of user’s confidence, and filtered messages by SMS center according to user’s confidence. Implemented three kinds of filtering technology (black/white list based, key words based, content based) on junk short message filtering method, which increase the efficiency very significantly.
Key words：junk shortmessage; user’s confidence; sample filtering; text categorization
近几年来，由于移动通信技术的快速发展，催化了诸多增值服务的产生。（剩余1796字）