Repositorio Digital de la UPF
Publicación
Contribución de congreso (ConferenceOutput)
Reference bias in monolingual machine translation evaluation
Comunicació presentada a la 54th Annual Meeting of the Association for Computational Linguistics, celebrada del 7 al 10 d'agost de 2016 a Berlín, Alemanya., In the translation industry, human translations are assessed by comparison with the source texts. In the Machine Translation (MT) research community, however, it is a common practice to perform quality assessment using a reference translation instead of the source text. In this paper we show that this practice has a serious issue – annotators are strongly biased by the reference translation provided, and this can have a negative impact on the assessment of MT quality., Marina Fomicheva was supported by funding from IULA (UPF) and the FI-DGR grant program of the Generalitat de Catalunya. Lucia Specia was supported by the QT21 project (H2020 No. 645452).
Proyecto:
EC / H2020 645452