QE Reading List

A paper list for machine translation quality estimation.

Welcome to add and criticism.

Model Architecture

• [EMNLP 2021] Levenshtein Training for Word-level Quality Estimation [paper]

• [COLING 2020] TransQuest: Translation Quality Estimation with Cross-lingual Transformers [paper] [code] [doc]

• [WMT 2019] QE BERT: Bilingual BERT Using Multi-task Learning for Neural Quality Estimation [paper]

• [AAAI 2019] “Bilingual Expert” Can Find Translation Errors [paper]

• [WMT 2017] Predictor-Estimator using Multilevel Task Learning with Stack Propagation for Neural Quality Estimation [paper]

Analysis

• [ACL 2020] Are we Estimating or Guesstimating Translation Quality? [paper]

Data Augmentation

• [AAAI 2021] DirectQE: Direct Pretraining for Machine Translation Quality Estimation [paper]

• [WMT 2020] Two-Phase Cross-Lingual Language Model Fine-Tuning for Machine Translation Quality Estimation [paper]

Multilingual QE

• [ACL 2021] An Exploratory Analysis of Multilingual Word-Level Quality Estimation with Cross-Lingual Transformers [paper]

• [AACL 2020] An Exploratory Study on Multilingual Quality Estimation [paper]

Multi-task QE

• [ACL Findings 2022] Better Quality Estimation for Low Resource Corpus Mining [paper]

• [EMNLP 2020] Translation Quality Estimation by Jointly Learning to Score and Rank [paper]

Self-supervised & Unsuperised QE

• [ACL 2022] As Little as Possible, as Much as Necessary: Detecting Over- and Undertranslations with Contrastive Conditioning [paper] [code]

• [EMNLP 2021] Self-Supervised Quality Estimation for Machine Translation [paper] [code]

• [COLING 2020] Intermediate Self-supervised Learning for Machine Translation Quality Estimation [paper]

• [TACL 2020] Unsupervised Quality Estimation for Neural Machine Translation [paper]

Glass-box

• [WMT 2020] IST-Unbabel Participation in the WMT20 Quality Estimation Shared Task [paper]

Efficiency

• [EMNLP 2021] Classification-based Quality Estimation: Small and Efficient Models for Real-world Applications [paper]

Knowledge Distillation QE

• [ACL Findings 2021] Knowledge Distillation for Quality Estimation [paper]

QE for Image Captions

• [NAACL 2021] Quality Estimation for Image Captions Based on Large-scale Human Evaluations [paper]

WMT QE

WMT QE Reports

• [WMT 2020] Findings of the WMT 2020 Shared Task on Quality Estimation [paper]

• [WMT 2019] Findings of the WMT 2019 Shared Tasks on Quality Estimation [paper]

WMT QE Winners

WMT 2021

• [WMT 2021] QEMind: Alibaba's Submission to the WMT21 Quality Estimation Shared Task [paper]

WMT 2020

• [COLING 2020] TransQuest: Translation Quality Estimation with Cross-lingual Transformers [paper] [code] [doc]

• [WMT 2020] TransQuest at WMT2020: Sentence-Level Direct Assessment [paper]

WMT 2019

• [WMT 2019] Unbabel’s Participation in the WMT19 Translation Quality Estimation Shared Task [paper]

WMT 2018

• [AAAI 2019] “Bilingual Expert” Can Find Translation Errors [paper]

• [WMT 2018] Alibaba Submission for WMT18 Quality Estimation Task [paper]

WMT 2017

• [WMT 2017] Predictor-Estimator using Multilevel Task Learning with Stack Propagation for Neural Quality Estimation [paper]

• [ACM 2017] Predictor-estimator: Neural quality estimation based on target word prediction for machine translation

(Waiting for Classification)

• [ACL 2022] Bias Mitigation in Machine Translation Quality Estimation

• [EMNLP 2021] Investigating the Helpfulness of Word-Level Quality Estimation for Post-Editing Machine Translation Output

• [EMNLP 2021] deepQuest-py: Large and Distilled Models for Quality Estimation

• [EMNLP Findings 2021] Beyond Glass-Box Features: Uncertainty Quantification Enhanced Quality Estimation for Neural Machine Translation [paper]

• [EMNLP Workshop 2021] The Eval4NLP Shared Task on Explainable Quality Estimation: Overview and Results [paper]

• [WMT 2021] Ensemble Fine-tuned mBERT for Translation Quality Estimation [paper]

• [WMT 2021] Pushing the Right Buttons: Adversarial Evaluation of Quality Estimation [paper]

• [WMT 2021] The JHU-Microsoft Submission for WMT21 Quality Estimation Shared Task [paper]

• [WMT 2019] YiSi - a Unified Semantic MT Quality Evaluation and Estimation Metric for Languages with Different Levels of Available Resources [paper] [code]

• [WMT 2019] Quality Estimation and Translation Metrics via Pre-trained Word and Sentence Embeddings [paper]