Anazawa Ryoko, Ishikawa Hirono, Takahiro Kiuchi
Graduate School of Medicine, The University of Tokyo, Tokyo, Japan.
Comput Inform Nurs. 2013 Aug;31(8):382-7. doi: 10.1097/NXN.0b013e3182999dc2.
Nurses whose native language is not English may face a language barrier when they read nursing literature published internationally. Freely accessible online machine translation could be one possible solution to reduce their burden. In the present study, English-Japanese translations of nursing article abstracts from the major online machine translation system, Google Translate, were evaluated by 250 nursing faculty members across Japan using a previously verified method for evaluating intelligibility. The abstracts were evaluated according to sections typically classified in structured abstracts: title, background, aim, method, result, discussion, conclusion, and relevance to clinical practice. We also asked for respondents' impressions on the perceived usefulness of the translated abstract as a whole. The results indicated that Google Translate had minimally acceptable quality and identified the word count and the existence of unidentifiable words of source texts as factors contributing to translation quality. Nurses may benefit from Google Translate when reading the professional literature. However, our results suggest that more effort should be made to improve online machine translation performance, and emphasize the importance of providing training and education for nursing professionals to utilize this technology.
母语非英语的护士在阅读国际出版的护理文献时可能会面临语言障碍。免费在线机器翻译可能是减轻他们负担的一种解决方案。在本研究中，日本各地的250名护理教员使用先前验证过的可理解性评估方法，对主要在线机器翻译系统谷歌翻译的护理文章摘要的英日译文进行了评估。摘要根据结构化摘要中通常分类的部分进行评估：标题、背景、目的、方法、结果、讨论、结论以及与临床实践的相关性。我们还询问了受访者对翻译后的摘要整体感知有用性的印象。结果表明，谷歌翻译的质量勉强可以接受，并确定源文本的字数和存在无法识别的单词是影响翻译质量的因素。护士在阅读专业文献时可能会从谷歌翻译中受益。然而，我们的结果表明，应做出更多努力来提高在线机器翻译性能，并强调为护理专业人员提供利用这项技术的培训和教育的重要性。
