科技术语汉英机器翻译质量研究——以测绘学术语为例
Exploring the Chinese-English Machine Translation Quality of Technical Terms—A Case Study of Terms in Surveying Sciences

本研究选取谷歌翻译这一主流机器翻译引擎，以测绘学术语为例，通过翻译实验考察了科技术语汉英机器翻译的质量。研究发现：所选术语的机器翻译准确率达到七成，表明机器翻译确有相当的应用价值。研究重点考查了科技术语汉译英机器翻译中主要存在的五种翻译错误，分别为术语含义判断错误、译文近义词混用、欠译、词性判断错误以及译文不合行业规范，并结合实例进行了详细分析。本研究在一定程度上揭示了现有主流机器翻译引擎在科技术语汉英翻译方面的整体翻译水平和问题所在，能够在译前编辑和译后编辑方面为机器翻译的广大用户提供一定参考。
Employing Google Translation, one of the mainstream Machine Translation (MT) engines, and taking terms in surveying sciences as an example, translation experiments were conducted in this study to examine the quality of target texts in the Chinese-English machine translation of technical terms. It is found that: around 70 percent of the terms we selected were accurately translated into English, indicating the considerable value of MT in practice. Particularly, this study focuses on the five major types of translation errors in the Chinese-English machine translation of technical terms as revealed in the translation experiments, including mis-decoding the meaning of the source terms, misuse of synonyms, inadequate translation, mis-decoding of part of speech and translated texts inconsistent with traditions of a specific field, and detailed exposition are also provided in the form of case analysis. This study is meaningful in that, it reveals the merits and demerits of current mainstream MT engines in terms of C-E translation of technical terms. Hopefully, it will shed some light on such aspects as pre-editing and post-editing references for users of MT.