机器翻译的译前与译后编辑在科技文本翻译中的探究

Research on the Pre-edit and Post-edit of Machine Translation in Science and Technology Text Translation

随着社会对科技文本高负荷的翻译需求和应用型翻译人才培养需要,基于大数据的互联网翻译成为机器翻译技术实用化的突破口。本研究以非英语专业的大二学生三组共30人为对象,以科技文章中的中文译文为测试原材料,采用互联网大数据机器翻译和BLAST平行语料分析软件,整合翻译专业教师评价和语言错误分析两种评估模式,考察译前和译后编辑加工在提升双语文本信息转换质量方面的作用,探索人助机译的翻译教学路径。结果表明,在科技文本翻译过程中,增加译前与译后编辑指导的互联网机器翻译完成质量优于传统人工翻译,同时也优于仅在机器翻译后进行编辑加工的翻译任务。

With the urgent need of science and technology text translation and applied translators, the big-data-based Internet translation represents a breakthrough in machine translation technology. Based on the comparison of three tasks with ten second grade non-English major college students in each group, with the help of the internet big data translation and BLAST parallel corpus analyzing software, and integrating the e- valuation of machine error analysis and professional translation teachers, this article intends to study how to improve the translation quality from the perspective of pre-edit and post-edit,to explore the potential teaching path with the big-data-based Internet translation and human editing. The result shows that,in the process of scientific text translation,the translation task,containing the pre-edit task of language processing comparison between natural language and the internet machine translation and the task of post-edit,is more efficient than the traditional human translation,and also than the internet translation plus post-edit task.