MLQE-PE: A Multilingual Quality Estimation and Post-Editing Dataset
Authors:
- Marina FomichevaMFMarina Fomicheva
- Shuo SunSSShuo Sun
- Erick FonsecaEFErick Fonseca
- Chrysoula ZervaCZChrysoula Zerva
- Frédéric BlainFBFrédéric Blain
- Vishrav ChaudharyVCVishrav Chaudhary
- Francisco GuzmánFGFrancisco Guzmán
- Nina LopatinaNLNina Lopatina
- Lucia SpeciaLSLucia Specia
- André F. T. MartinsAMAndré F. T. Martins
Preprint
Abstract
We present MLQE-PE, a new dataset for Machine Translation (MT) Quality Estimation (QE) and Automatic Post-Editing (APE). The dataset contains eleven language pairs, with human labels for up to 10,000 translations per language pair in the following formats: sentence-level direct assessments and post-editing effort, and word-level good/bad labels. It also contains the post-edited sentences, as well as titles of the articles where the sentences were extracted from, and the neural MT models used to translate the text.Keywords
All Related Versions
This publication has 4 references indexed in Scilit:
- WikiMatrix: Mining 135M Parallel Sentences in 1620 Language Pairs from Wikipedia2021
- Unsupervised Cross-lingual Representation Learning at Scale2020
- Findings of the WMT 2019 Shared Tasks on Quality Estimation2019
- Quality Estimation for Machine TranslationSynthesis Lectures on Human Language Technologies, 2018