Despite the importance of trust in any work environment, this concept has rarely been investigated for MT. The present contribution aims at filling this gap by presenting a post-editing experiment carried out with translator trainees. An institutional academic text was translated from Italian into English. All participants worked on the same target text. Half of them were told that the text was a human translation needing revision, while the other half was told that it was an MT output to be postedited. Temporal and technical effort were measured based on words per second and HTER. Results were complemented with a manual analysis of a subset of the observations.

Do translator trainees trust machine translation? An experiment on post-editing and revision

Luisa Bentivogli
2019

Abstract

Despite the importance of trust in any work environment, this concept has rarely been investigated for MT. The present contribution aims at filling this gap by presenting a post-editing experiment carried out with translator trainees. An institutional academic text was translated from Italian into English. All participants worked on the same target text. Half of them were told that the text was a human translation needing revision, while the other half was told that it was an MT output to be postedited. Temporal and technical effort were measured based on words per second and HTER. Results were complemented with a manual analysis of a subset of the observations.
File in questo prodotto:
File Dimensione Formato  
MT-Summit2019-translators.pdf

accesso aperto

Descrizione: Articolo principale
Licenza: Creative commons
Dimensione 403.1 kB
Formato Adobe PDF
403.1 kB Adobe PDF Visualizza/Apri

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: http://hdl.handle.net/11582/320664
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
social impact