Automatic post-editing of machine translation: a neural programmer-interpreter approach

Thuy-Trang Vu, Reza Haffari

Research output: Chapter in Book/Report/Conference proceedingConference PaperResearchpeer-review

2 Citations (Scopus)


Automated Post-Editing (PE) is the task of automatically correcting common and repetitive errors found in machine translation (MT) output. In this paper, we present a neural programmer-interpreter approach to this task, resembling the way that humans perform post-editing using discrete edit operations, which we refer to as programs. Our model outperforms previous neural models for inducing PE programs on the WMT17 APE task for German-English up to +1 BLEU score and - 0.7 TER scores.

Original languageEnglish
Title of host publicationEMNLP 2018
Subtitle of host publicationBrussels, Belgium Oct. 31-Nov. 4
EditorsDavid Chiang, Julia Hockenmaier, Jun'ichi Tsujii
Place of PublicationStroudsburg PA USA
PublisherAssociation for Computational Linguistics (ACL)
Number of pages6
ISBN (Electronic)9781948087841
Publication statusPublished - 2018
EventEmpirical Methods in Natural Language Processing 2018 - Brussels, Belgium
Duration: 31 Oct 20184 Nov 2018 (Proceedings)


ConferenceEmpirical Methods in Natural Language Processing 2018
Abbreviated titleEMNLP 2018
Internet address

Cite this