三款机器翻译译后编辑工具对比研究

A Comparative Study of Three Machine Translation Post-editing Tools

随着人工智能神经网络翻译质量的日益提升,语言服务企业和翻译研究学界对机器翻译+译后编辑(MTPE)模式的接受度越来越高,并从各个角度不断探索提高MTPE质量与效率的方法策略。其中,工具平台的使用直接影响译后编辑的质量和效率。基于此,选择YiCAT、LanguageX和GT4T三款同时聚合多个机器翻译引擎的平台工具,结合民航《风险管理手册》翻译实践,从项目创建、编辑页面布局和语料管理3个维度分析其异同与优劣,探索工具选择时应考虑的各方因素以及MTPE平台未来发展的趋势。

As the quality of AI-based neural machine translation keeps improving,both language service companies and translation research academia are increasingly embracing the idea of Machine Translation Post-Editing or MTPE.Meanwhile,methods to improve the quality and efficiency of MTPE have been explored from multiple perspectives.As the selection of an appropriate platform plays a key role in this regard,it is worth further study.This paper has chosen YiCAT、LanguageX and GT4T as the research subjects and tries to compare the three as they are being applied to the MTPE of Risk Management Handbook.A detailed analysis of the three platforms,which have all integrated several machine translation engines,is conducted from the angels of project setup,editing process and language asset management.Finally,conclusions are drawn regarding factors to be considered for choosing MTPE tools as well as the development trends of such platforms.