学生译员使用IFLYREC对其英中同传口译质量影响研究

近年来人工智能飞速发展,并在众多领域得到应用。在口笔译领域,语音翻译器正在兴起和推广。文章旨在研究在同传口译中,当学生译员遇到术语、长句和信息密集句等口译难点时,以科大讯飞IFLYREC为代表的语音翻译器能否帮助他们提升口译质量。实验数据和问卷结果显示,IFLYREC有利于提升学生译员的术语口译表现,但对长句和信息密集句,IFLYREC不一定能帮助他们提升口译表现。因此,为使IFLYREC更好地辅助人类译员,建议研发者做如下改进:减少实时翻译的延迟,提高译文准确性,优化语法纠错功能。

In recent years, Artificial Intelligence(AI) has been developing very fast and found application in many fields. In the field of translation and interpreting, more and more Speech Translation(ST) tools have been introduced to assist human interpreters and audience. This thesis aims to find out whether IFLYREC can help trainee interpreters improve their simultaneous interpreting performance when they encounter “problem triggers”. The experiment results and questionnaire data show that in general, IFLYREC helped improve trainee interpreters’ terminology performance in SI, but their performance on long sentences and sentences with condensed information were not necessarily improved with the use of IFLYREC. Thus, in order for IFLYREC to better assist human interpreters, the developers should consider making certain improvements, including reducing the delay of the real-time translation by IFLYREC, improving its translation accuracy and optimizing its grammatical error correction system.