摘要
本文提出了一种适用于对互联网发送的垃圾短信进行监控和过滤的解决方案,该方案根据短信日发送频度、短信内容、短信长度等特征,采用模糊匹配和频度监控相结合的方法进行短信监控和过滤。系统中创新性地提出了内容预变换、关键字加权控制频度等方法,相比传统方法而言,本系统在过滤准确度和误判率两方面的指标均获得大幅度提升。
This paper presents an approach to monitor and filter the spam SMS originated from the Internet effectively. Our approach provides fuzzy matching and frequency monitoring algorithms to filter the spam short messages with the characteristics of message contents and length. We introduce two methods, the content transformation and
weight based frequency monitoring, which can improve the accuracy of spam short message filtering and decrease the mis-decision ratio comparing with the traditional spam short message filtering methods.
出处
《电信科学》
北大核心
2008年第5期61-67,共7页
Telecommunications Science
关键词
垃圾短信过滤
模糊匹配
频度监控
spam short message filtering, fuzzy matching, frequency monitoring
