Explainable Quality Estimation: CUNI Eval4NLP Submission

Peter Polák, Muskaan Singh, Ondřej Bojar


Abstract
This paper describes our participating system in the shared task Explainable quality estimation of 2nd Workshop on Evaluation & Comparison of NLP Systems. The task of quality estimation (QE, a.k.a. reference-free evaluation) is to predict the quality of MT output at inference time without access to reference translations. In this proposed work, we first build a word-level quality estimation model, then we finetune this model for sentence-level QE. Our proposed models achieve near state-of-the-art results. In the word-level QE, we place 2nd and 3rd on the supervised Ro-En and Et-En test sets. In the sentence-level QE, we achieve a relative improvement of 8.86% (Ro-En) and 10.6% (Et-En) in terms of the Pearson correlation coefficient over the baseline model.
Anthology ID:
2021.eval4nlp-1.24
Volume:
Proceedings of the 2nd Workshop on Evaluation and Comparison of NLP Systems
Month:
November
Year:
2021
Address:
Punta Cana, Dominican Republic
Editors:
Yang Gao, Steffen Eger, Wei Zhao, Piyawat Lertvittayakumjorn, Marina Fomicheva
Venue:
Eval4NLP
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
250–255
Language:
URL:
https://aclanthology.org/2021.eval4nlp-1.24
DOI:
10.18653/v1/2021.eval4nlp-1.24
Bibkey:
Cite (ACL):
Peter Polák, Muskaan Singh, and Ondřej Bojar. 2021. Explainable Quality Estimation: CUNI Eval4NLP Submission. In Proceedings of the 2nd Workshop on Evaluation and Comparison of NLP Systems, pages 250–255, Punta Cana, Dominican Republic. Association for Computational Linguistics.
Cite (Informal):
Explainable Quality Estimation: CUNI Eval4NLP Submission (Polák et al., Eval4NLP 2021)
Copy Citation:
PDF:
https://preview.aclanthology.org/nschneid-patch-2/2021.eval4nlp-1.24.pdf
Code
 pe-trik/eval4nlp-2021