Communicating the environmental crisis to non-specialist audiences : A quality assessment of machine translated subtitles into three different languages
McDonagh, Sarah (Universitat Autònoma de Barcelona)
Pujadas Farreras, Marina (Universitat Autònoma de Barcelona)
Barrosso Ordóñez, Martín (PANGEANIC)

Date:	2023
Abstract:	In recent years, the language and terms used to describe the climate crisis have increased in urgency in response to the worsening ecological state of the natural world (Carrington 2019; Vinter 2021). This has coincided with an increase in environmental information more generally with agencies and organisations sharing up-to-date ecological data with policy makers and the public. Given the large quantity of environmental information now available, its translation into different languages can be costly and time-consuming. Machine translation offers a way in which to support the process of providing up-to-date environmental information in a range of different languages. This presentation will report on the results of an exploratory study using machine translation, provided by the language technological provider Pangeanic, of an informational video about the GreenSCENT project. GreenSCENT (Smart Citizen Engagement for a Green Future) (Nº 101036480) is a three-year funded European H2020 project that aims to foster and promote positive behavioural change towards the environment through the development of educational materials and tools (https://www. green-scent. eu/). This study will examine the accuracy of machine translated subtitles from English into three different languages (Italian, Spanish and Catalan) of the GreenSCENT introductory video (2022). Following the Bilingual Evaluation Understudy (BLEU) metric and the more recent Metric for Evaluation of Translation with Explicit ORdering (METEOR), quality will be assessed by comparing the machine translated subtitles in each language with those created by a human.
Grants:	European Commission. Horizon 2020 101036480
Rights:	Aquest document està subjecte a una llicència d'ús Creative Commons en que s'entrega l'obra al domini públic. Podeu copiar, modificar, distribuir l'obra i fer-ne comunicació pública, fins i tot amb finalitat comercial, sense demanar cap mena de permís.
Language:	Anglès
Series:	GreenSCENT
Document:	Contribució a congrés
Subject:	Climate change ; Machine translation ; Sustainability ; GreenSCENT project ; Quality
Published in:	3rd Wrocław Terminological Meeting TERMOS 2023. Wrocław, 3 : 2023

22 p, 1.6 MB

The record appears in these collections:
Contributions to meetings and congresses > Presentations