一种基于特征值的短信过滤匹配算法  

Short Message Filtering and Matching Algorithm Based on Eigenvalues

作　　者：钱苏林[1,2] 李炜[1,2] 王晶[1,2] 

机构地区：[1]北京邮电大学网络与交换技术国家重点实验室,北京100876 [2]东信北邮信息技术有限公司,北京100191

出　　处：《计算机系统应用》2012年第5期55-62,共8页Computer Systems & Applications

基　　金：国家自然科学基金(61072057;60902051);国家973计划(2012CB315802);中央高校基本科研业务费专项资金(BUPT2009RC0505);国家科技重大专项(2011ZX03002-001-01;2011ZX03002-002-01)

摘　　要：短信营销业务是近年来发展较快的短信业务之一,而随着该项业务的广泛应用,对短信的过滤匹配功能也提出了更高的要求。结合短信营销业务的特性,提出了一种高效的短信过滤匹配算法。算法依据短信分片特征值进行短信过滤,采用位向量法、编辑距离算法进行短信匹配,实现了高效的海量短信过滤匹配功能。文中给出了特征值算法的详细步骤,并对算法的有效性和正确性进行了分析对比。SMS marketing is growing at a fast rate in the past few years.As this business is widely used,the requirement for the filtering and matching function of the short message is becoming increasingly higher.Having considering some characteristics of SMS advertising,we present an algorithm for effective text filtering and matching in this article.More specifically,the algorithm is briefly described as followed： firstly,text is filtered based on its eigenvalue and secondly using bit-vector method and Levenshtein Distance algorithm to match two candidate short messages thereby realizing the efficiency of mass short message filtering and matching function.This paper describes the details on calculation of text eigenvalue and presents the analysis on the effectiveness and correctness of this algorithm.

关 键 词：短信 特征值 编辑距离 高效过滤匹配 

分 类 号：TN929.53[电子电信—通信与信息系统]