Machine Translation Post-Editing Levels: Breaking Away from the Tradition and Delivering a Tailored Service
2948Authors
nan
Abstract
AbstractWhile definitions of full and light post-editing have been around for a while, and error typologies like DQF and MQM gained in prominence since the beginning of last decade, for a long time customers tended to refuse to be flexible as for their final quality requirements, irrespective of the text type, purpose, target audience etc. We are now finally seeing some change in this space, with a renewed interest in different machine translation (MT) and post-editing (PE) service levels. While existing definitions of light and full post-editing are useful as general guidelines, they typically remain too abstract and inflexible both for translation buyers and linguists. Besides, they are inconsistent and overlap across the literature and different Language Service Providers (LSPs). In this paper, we comment on existing industry standards and share our experience on several challenges, as well as ways to steer customer conversations and provide clear instructions to post-editors.
Keywords
Full Content
Machine Translation Post-Editing Levels: Breaking Away from the Tradition and Delivering a Tailored Service Mara Nunziatini Welocalize Italy S.r.l. Via Alserio 22, Milan, Italy mara.nunziatini@welocalize.com Lena Marg Welocalize Inc. Frederick, MD, United States lena.marg@welocalize.com Abstract While definitions of full and light post- editing have been around for a while, and error typologies like DQF and MQM gained in prominence since the beginning of last decade, for a long time customers tended to refuse to be flexible as for their final quality requirements, irrespective of the text type, purpose, target audience etc. We are now finally seeing some change in this space, with a renewed in- terest in different machine translation (MT) and post-editing (PE) service lev- els. While existing definitions of light and full post-editing are useful as general guidelines, they typically remain too ab- stract and inflexible both for translation buyers and linguists. Besides, they are inconsistent and overlap across the litera- ture and different Language Service Pro- viders (LSPs). In this paper, we would like to comment on existing industry standards and share our experience on several challenges, as well as ways to steer customer conversations and provide clear instructions to post-editors. 1 Introduction As one of the largest multilingual LSPs, we have been offering machine translation post-editing services for many years, and our team supports more than 30 of our largest customers in the Enterprise or Regulated space with MT and post- editing programs in often 30+ language pairs. When implementing machine translation for a new customer, we always provide a post-editing training to the linguists working on the program. © 2020 The authors. This article is licensed under a Creative Commons 3.0 licence, no derivative works, at- tribution, CC-BY-ND. During this training, among other relevant topics, we focus on the basics of post-editing and we explain what the client’s requirements are regarding final translation quality. Based on experience, we found that it can be very complicated to communicate what is expected of linguists in different post-editing levels. While it is easier to explain what is expected of light versus full post-editing, there are some grey areas that don’t fall either into the full post-editing or the light post-editing service. Furthermore, our customers will often not be experts of translation quality assurance methodologies, and also not be familiar with the common definitions of the different levels of post-editing. As such, they are themselves often not entirely sure which approach would meet, exceed or fall short of their requirements. It is therefore crucial to guide them and define their requirements from the very outset, also in order to be able to clearly communicate them to the post-editors. This is extremely important since post-editors might feel confused if they do not receive clear instructions, and will probably end up delivering a quality that is either too high – in this case they will not be productive – or too low – and the clients’ quality requirements will not be met. Ultimately, the effort of the post-editor depends strictly on …
Content truncated to 500 words
Labeled Text
AbstractWhile definitions of full and light post-editing have been around for a while, and error typologies like DQF and MQM gained in prominence since the beginning of last decade, for a long time customers tended to refuse to be flexible as for their final quality requirements, irrespective of the text type, purpose, target audience etc. We are now finally seeing some change in this space, with a renewed interest in different machine translation (MT) and post-editing (PE) service levels. While existing definitions of light and full post-editing are useful as general guidelines, they typically remain too abstract and inflexible both for translation buyers and linguists. Besides, they are inconsistent and overlap across the literature and different Language Service Providers (LSPs). In this paper, we comment on existing industry standards and share our experience on several challenges, as well as ways to steer customer conversations and provide clear instructions to post-editors.
Publication Details
- Paper ID:
- 2948
- Year:
- 2020
- Publication:
- TIPSTER
- Publication Type:
- conf
- DOI:
- nan