基于朴素贝叶斯算法的语言特征描述研究  

Research of Linguistic Features Based on Naive Bayes

作　　者：李森[1] 赵洁[1] 

机构地区：[1]北方工业大学,北京100144

出　　处：《农业网络信息》2012年第5期34-36,共3页Agriculture Network Information

摘　　要：近年来计算语言学理论与方法的迅速发展与逐渐成熟带动了文本挖掘等交叉领域的进步,取得了一定的实践成果,但通过交叉领域的进步反过来促进语言学发展的研究却相对较少。本研究从网络新闻语料入手,基于朴素贝叶斯算法,探究了特征词数与真实语料的关系,为理论与应用研究提供了相关数据与思路。In recent years,the rapid development of computational linguistics theory and the gradual maturity of its method led to the improvements in the cross field such as text mining and certain achievements were obtained.On the contrary,however,there was relatively little study on the promotion of cross field achievement on the development of linguistics research.In this paper,based on the network news corpus and the Naive Bayes,the relationship between the feature words and real corpu was discussed,providing correlated data and thoughts for theory and applied research.

关 键 词：特征选择 NAIVE BAYES 朴素贝叶斯 文本分类 语言模型 

分 类 号：TP315[自动化与计算机技术—计算机软件与理论]