基于Hash函数和贝叶斯方法的垃圾短信在线过滤系统  

Spam message online filtering system based on hash function and naive Bayes

作　　者：范晶[1] 刘菊新[1] 陈峰[2] 徐从富[1] 

机构地区：[1]浙江大学计算机科学与技术学院,杭州310027 [2]浙江省公安厅科技处,杭州310009

出　　处：《计算机应用》2008年第4期1084-1086,共3页journal of Computer Applications

基　　金：国家863计划项目(2007AA01Z197);国家自然科学基金资助项目(60402010)

摘　　要：提出了一种综合集成黑白名单过滤模块和综合评价函数模块的垃圾短信在线过滤系统架构,引入了两层Hash函数和基于朴素贝叶斯的概率计算方法,对疑似垃圾短信从内容、长度、频率等特征上进行深入分析,实现了海量短信的实时高效过滤,并有效地解决了传统垃圾短信过滤系统中存在的问题。A spam message online filtering system was proposed, integrating the black-white list filtering module and synthesis value function module. A two-layer hash method and a probability computing method based on naive Bayes were used to analyze the suspect spam message in depth in terms of content, length, frequency and others. Excellent real-time and features in filtering the great many of short messages were achieved, and the problems of traditional spam message filtering system was resolved efficiently.

关 键 词：垃圾短信过滤 HASH函数 朴素贝叶斯方法 

分 类 号：TP319[自动化与计算机技术—计算机软件与理论] TP393.09[自动化与计算机技术—计算机科学与技术]