口译语料库的编码与标记:以SIDB为例

Encoding and tagging of interpretation corpus: Taking SIDB as an example

关于口译语料库的建置与研究,迄今以名古屋大学1999-2003年之间开发的英日/日英同步口译语料库(SIDB,Simultaneous Interpretation Data Base)为最大,约达100万字。该研究团队在此基础上,进行了一系列检证口译产出特征的量化研究。本研究首先将针对SIDB口译语料库加以评估,并分析其编码与标记之功能及特征。其次,亦将评述前述口译语料库所展开的研究成果,以厘清该语料库之定位及效益。借此,期能将现行语料之编码与标记具有之优势及其应避免之劣势,提出客观描述与评估。

Between 1999 and 2003,Nagoya University created a simultaneous interpretation database( SIDB) for English-Japanese and Japanese-English simultaneous interpretation. To date,this is the largest such database in the world,with around one million words. The research team at Nagoya conducted a series of quantitative studies on this database,to examine and confirm the characteristics of interpretation output. This study evaluates the SIDB corpus,and analyzes the qualities of its encoding and labeling. Additionally,this study evaluates and describes the results of the research into the SIDB,to establish the role and advantages of this system. Through this,it is possible to establish objective descriptions and evaluations of the benefits and pitfalls of the encoding and labeling of current corpora.