摘要
对Google、Systran和Bing三个在线翻译系统在德州历史数字图书馆元数据翻译性能上的表现进行人工评价,评价指标包括:流利程度、充分程度、误译数目和漏译数目。分析得出Google和Bing在没有经过任何语料训练的情况下均达到或超越"非母语中文"的翻译水平,最后提出利用在线翻译系统实现数字图书馆多语言信息存取的几种策略。
In this paper, performance of online translation systems including Google, Systran and Bing on translating metadata re- cords derived from a digital library- Portal to Texas History- is manually evaluated using four measures： fluency, adequacy, incorrect translation, and missing translation. The results demonstrated that Google and Bing could achieve ＂non-native Chinese＂ level of per- formance. The paper concludes with three possible strategies of implementing muhilingual information access in digital libraries applying machine translation to metadata records.
出处
《图书情报工作》
CSSCI
北大核心
2011年第2期16-20,111,共6页
Library and Information Service
