名词团块分裂移位的计算机处理研究

Computer processing research of phenomenon of Noun phrase splitting and movement

在自然语言的语句中,名词团块的分裂现象十分常见,且伴随有移位的出现,对计算机处理及机器翻译准确率的影响极大。HNC理论借鉴现代汉语语法学对名词性词组分裂移位的研究成果,采用以主辅变换为核心的处理方式对这类现象进行处理,最终形成相关的处理规则,使得计算机可通过这类规则自动识别可分裂移位的语句,提高机器翻译的准确率。从计算语言学研究处理的角度,详细介绍HNC理论主辅变换的研究成果,以及如何形成处理规则并供计算机使用,以期提高名词团块分裂的机器识别及处理的准确率。

The Phenomenon of Noun Phrase Splitting and Movement（PNPSM） can be easily observed in nature language.The problem influences seriously the accuracy rate of the natural language understanding and the machine translation system.The HNC theory carries on processing with taking the main-auxiliary chunk transformation as the core processing way to this kind of phenomenon,and finally forms the related processing rules,and causes the computer to automatic diagnosis the sentence through the rules,and enhances the accuracy rate of the machine translation system.The article introduces the research results of the main-auxiliary chunk transformation,and how does HNC form the processing rules and supplies them to computer,from computational linguistics angle ,in order to enhance the accuracy rate of the machine translation system.