Slator Machine Translation Expert-in-the-Loop Report

WakeSpace Repository

Show simple item record

dc.date.accessioned 2022-10-20T19:02:00Z
dc.date.available 2022-10-20T19:02:00Z
dc.date.issued 2022
dc.identifier.uri http://hdl.handle.net/10339/101517
dc.description.abstract The Slator Machine Translation Expert-in-theLoop Report provides a comprehensive view on the interaction between human experts and machines in translation production. In this report, we investigate the different ways that machines and human experts interact throughout the translation process, including project management, translation, and vendor selection. The report first examines the concept of humanmachine interaction in translation, in which human experts and machines work together, taking on different roles with varying degrees of responsibility. We examine the latest advances shaping and defining the relationship between human experts and machines in translation production — how they impact skills, workflows, processes, and technology. We also provide insights into how to optimize and harmonize this relationship. Drawing on interviews with representatives from nine language service providers (LSPs), the report highlights how expert-in-the-loop translation (aka MT-enabled workflows) has already become the dominant translation method and continues to increase in adoption. A handy, one-page table, outlines the level of adoption of expert-in-the-loop translation workflows per vertical, provides examples of expert-in-the-loop translation in action for further reading, and identifies use cases for both non-MT and pure-MT workflows. The report describes the machines that are involved in translation production as well as the role and characteristics of the human expert, which is typically (but not always) the post-editor. The most substantial chapter of the report, Expert-in-the-Loop Translation Production, examines how AI can be applied to support not only translators within the translation / postediting task but also project managers (PMs) in managing expert-in-the-loop translation workflows, selecting vendors, and more. The same chapter unpacks the different expertin-the-loop translation methods and explores the application and potential of tools, such as quality estimation (QE), automatic postediting, and quality assessment (QA) in expert-in-the-loop workflows, as well as the suitability of translator interfaces for expert-in-the-loop translation production. The final chapter, Expert-in-the-Loop Translation Pricing Models, outlines how technological advances brought about by MT, cloud computing, and AI-driven automation have affected (and continue to influence) industry pricing models. en_US
dc.language.iso en_US en_US
dc.publisher Slator en_US
dc.title Slator Machine Translation Expert-in-the-Loop Report en_US
dc.type Article en_US

Files in this item

This item appears in the following Collection(s)

Show simple item record