EM算法的研究及其在文本处理中的应用  

Research on EM Algorithm and Its Application in Text Processing

作　　者：安永辉[1] 

机构地区：[1]桂林电子科技大学计算机科学与工程学院,桂林541004

出　　处：《现代计算机》2013年第7期3-8,共6页Modern Computer

摘　　要：引入EM这一个经典的迭代算法,对EM算法进行具体的推导,并对算法的具体执行步聚做进一步的描述;接着对如今常用的几种文本分类的方法做了概括,重点分析Bayes分类方法,并将EM与Bayes分类方法进行结合,提高Bayes的分类精度,并在综述里对未来海量数据时代怎样来弥补EM算法的不足做简要的介绍。Introduces the current situation of the development of today, text processing, and introduces EM iterative algorithm for the classic one, and the derivation of the EM algorithm, and the specific implementation steps of algorithm of poly a further description; then after the summary of sev- eral methods commonly used nowadays in text classification, focuses on the analysis of Bayes classification methods, and the combination of EM and Bayes algorithm, application of en- hanced Bayes classifier in text processing and text classification, indirectly, to improve the classification accuracy of Bayes, and describes the deficiencies in the face of the future data times how to compensate for the EM algorithm.

关 键 词：EM BAYES分类器 文本处理 文本分类 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术]