摘要
机译系统应该建立在具有语篇分析的大型语料库基础之上,语篇语料库内的原语和目的语文本不是单向、而应是双向或多向流动的网络系统。文本的译文也应该有历时和共时的区别,因为互文性已是翻译的某种本质特征。
A data\|driven Machine Translation system should be based on discourse analysis in a gigantic computer corpus, intertextualized diachronically and synchronically with all the information available from SL and TL, annotated with marks of register, situation types, function and meaning. The tagged information can be retrieved, further analyzed and added all the time, thus forming a dynamic model.
出处
《武汉大学学报（人文科学版）》
CSSCI
2003年第4期488-491,共4页
Wuhan University Journal (Humanity Sciences)
基金
广东省教育厅人文社会科学研究基金项目([2002]02SJD740002)
关键词
语篇分析
机器翻译
描写翻译学
语篇特征
discourse analysis
MT
descriptive translation studies
textuality
