With or without post-editing processes? Evidence for a gap in machine translation evaluation

no code implementations • WS 2019 • Caroline Rossi, Emmanuelle Esperan{\c{c}}a-Rodier

Machine Translation Translation

ACCOL\'E : Annotation Collaborative d'erreurs de traduction pour COrpus aLign\'Es (ACCOL\'E: A Collaborative Platform of Error Annotation for Aligned Corpus)

no code implementations • JEPTALNRECITAL 2018 • Francis Brunet-Manquat, Emmanuelle Esperan{\c{c}}a-Rodier

La plateforme ACCOL{\'E} (Annotation Collaborative d{'}erreurs de traduction pour COrpus aLign{\'E}s) propose une palette de services innovants permettant de r{\'e}pondre aux besoins modernes d{'}analyse d{'}erreurs de traduction : gestion simplifi{\'e}e des corpus et des typologies d{'}erreurs, annotation d{'}erreurs efficace, collaboration et/ou supervision lors de l{'}annotation, recherche de mod{\`e}le d{'}erreurs dans les annotations.

Collection of a Large Database of French-English SMT Output Corrections

no code implementations • LREC 2012 • Marion Potet, Emmanuelle Esperan{\c{c}}a-Rodier, Laurent Besacier, Herv{\'e} Blanchon

We also post-edited 1, 500 gold-standard reference translations (of bilingual parallel corpora generated by professional) and noticed that 72 {\%} of these translations needed to be corrected during post-edition.

Machine Translation Sentence +1