摘要
语言模型是对自然语言的描述,研完语言模型的构造方法是计算语言学的核心内容之一。本文在深入分析基于概率分布和上下文特征与信息的统计语言模型所依据的数学理论基础上,讨论了这两类语言模型的异同及其建立方法,并了讨论了基于组合思想的语言建模方法。
Language model is used to describe natural language,it is one of tasks of computational linguistics to study the ways of language modeling. This paper disscusses the model based on probability distribution and model based on context feature information, moreover ,the difference and modeling methods of these two models have presented.
出处
《计算机科学》
CSCD
北大核心
2004年第5期176-179,共4页
Computer Science
基金
山西省青年科技研究基金的资助(20021015)
