汉语句法分析中标点符号的运用  

The Use of Punctuations in Chinese Syntactic Parsing Techniques

作　　者：张小艳[1] 邵刚[1] 史月飞[1] 

机构地区：[1]西安科技大学计算机系,陕西西安710054

出　　处：《计算机工程与科学》2009年第1期145-147,共3页Computer Engineering & Science

基　　金：陕西省教育厅专项课题(06JK246)

摘　　要：目前,大部分句法分析都忽略标点符号这一重要的句法特征或者只进行非常简单的处理。本文根据标点符号的句法结构特性,提出规则分层的方法,将标点融入汉语句法分析中。利用标点符号的分割作用,将长句分成一个个小的句子的序列,并对每个小的句子单元进行句法和结构分析,再根据已经抽取出来的类型规则进行二次句法分析,从而得到一个完整的句法分析树。实验表明,这种方法不但解决了部分长句无法正确得到句法树的难题,而且分析的歧义减小了,效率得到了提高。So far, most Chinese syntactic parsing techniques neglect the punctuations or oversimplify their functions. However, it is actually very important information of syntactic characters. According to the features of punctuations in the syntactic structure, this paper proposes a new rule-layered approach. This method makes the punctuations into Chinese syntactic analysis and uses the punctuation role to split long sentences into small sequence sentences. Then each small unit is parsed syntactically and structurally. Finally, we extract the type rule to analyse and complete the parsing tree. Experiments show that this approach not only solves the problem that part of long sentences can not correctly obtain syntactic trees, but also reduces the ambiguities of parsing,and increases efficiency.

关 键 词：句法分析 标点符号 类型规则 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]