Pushing the Right Buttons:Adversarial Evaluation of Quality Estimation

Abstract

Current Machine Translation (MT) systems achieve very good results on a growing variety of language pairs and datasets. However, they are known to produce fluent translation outputs that can contain important meaning errors, thus undermining their reliability in practice. Quality Estimation (QE) is the task of automatically assessing the performance of MT systems at test time. Thus, in order to be useful, QE systems should be able to detect such errors. However, this ability is yet to be tested in the current evaluation practices, where QE systems are assessed only in terms of their correlation with human judgements. In this work, we bridge this gap by proposing a general methodology for adversarial testing of QE for MT. First, we show that despite a high correlation with human judgements achieved by the recent SOTA, certain types of meaning errors are still problematic for QE to detect. Second, we show that on average, the ability of a given model to discriminate between meaning-preserving and meaning-altering perturbations is predictive of its overall performance, thus potentially allowing for comparing QE systems without relying on manual quality annotation.

Publication DOI: https://doi.org/10.48550/arXiv.2109.10859
Additional Information: Copyright 2021 the authors, Creative Commons Attribution 4.0 International (CC BY 4.0)
Event Title: 6th Conference on Machine Translation, WMT 2021
Event Type: Other
Event Dates: 2021-11-10 - 2021-11-11
Uncontrolled Keywords: Language and Linguistics,Human-Computer Interaction,Software
ISBN: 9781954085947
Last Modified: 24 Apr 2024 17:51
Date Deposited: 01 Feb 2023 09:56
Full Text Link:
Related URLs: http://www.scop ... tnerID=8YFLogxK (Scopus URL)
https://arxiv.o ... /abs/2109.10859 (Publisher URL)
PURE Output Type: Conference contribution
Published Date: 2021-09-22
Authors: Kanojia, Diptesh
Fomicheva, Marina
Ranasinghe, Tharindu (ORCID Profile 0000-0003-3207-3821)
Blain, Frédéric
Orasan, Constantin
Specia, Lucia

Download

[img]

Version: Published Version

License: Creative Commons Attribution

| Preview

Export / Share Citation


Statistics

Additional statistics for this record