摘要
在系统功能语言学小句理论的基础上,通过对5021个经过自动句法分析的句子的主语、谓语、状语、补语、补语1/2/3、补语剩余部分和谓语剩余部分的识别错误进行了数据统计和分析。共得出3类一级错误、7类二级错误和38类三级错误,并且将错误成因细致分为4大类13小类,配有具体实例说明。对面向自然语言处理的功能句法分析和语料标注均有一定的实际意义。
Based on the Functional Clause theories,through the analysis of 5021 sentences after clause parsing experiments,data of the parsing errors were collected and analyzed. There were 193 kinds of errors in total,and they were classified into three categories： 3 kinds of first class errors,7 kinds of second class errors,and 38 kinds of third class errors. The cause of errors were also analyzed and divided into 4 categories with 13 kinds,and detailed examples were provided.
出处
《鸡西大学学报（综合版）》
2014年第9期124-127,共4页
JOurnal of Jixi University:comprehensive Edition
基金
大连理工大学2013年度教育部人文社会科学研究规划基金项目"机器翻译中的英语功能名词短语识别及汉译研究"(编号:13YJAZH062)
关键词
功能小句
自动句法分析
错误分析
functional clause
automatic parsing
error analysis
