摘要
以"汽车技术文献翻译语料库"为基础,对机器译文和人工译文进行了平行对比,描述了对机器译文中出现的名词短语、动词短语、介词短语、动词不定式短语、分词短语和从句等错译的标注和统计,归纳分析了各类句法错译的典型形式,缩小了对语言进行形式化处理的范围,为翻译系统在原有基础上补充建立形式化规则并最终提高机器翻译质量提供了语言分析基础。
This research marks the syntactic errors in the machine translation(MT) of noun phrases, verb phrases, prepositional phrases, infinitive phrases, participial phrases and various clauses on the basis of a sentence-level comparative study of MT and human translation from Automobile Document Translation Corpus. The paper details the statistic results of each type of errors and describes the typical patterns of syntactic errors in MT output, which aims to narrow down language rules in need of optimization or formalization in MT and finally improve the robustness of MT.
出处
《同济大学学报(社会科学版)》
CSSCI
北大核心
2014年第1期111-118,124,共9页
Journal of Tongji University:Social Science Edition
关键词
机器翻译
机器译文
人工译文
句法错误
machine translation (MT)
MT output
human translation
syntactic errors