一句话：提出结构不变性测试以验证机器翻译鲁棒性，弥补其难以评估的问题。
本文关注机器翻译系统，尤其是神经机器翻译（NMT）模型，其复杂性与不可解释性使鲁棒性难以评估，更难以保证。翻译系统可能返回劣质结果，进而导致误解、误诊、人身安全威胁或政治冲突。尽管机器翻译已广泛进入日常生活，验证其鲁棒性却非常困难，相关工作仍严重不足。为此，论文提出结构不变性测试（structure-invariant testing），通过检验译文在句子结构保持变换下是否保持一致/不变来发现系统缺陷，为机器翻译鲁棒性验证提供一种测试思路。摘要主要阐述问题动机与必要性，未给出具体实验数据。
In recent years, machine translation software has increasingly been integrated into our daily lives. People routinely use machine translation for various applications, such as describing symptoms to a foreign doctor and reading political news in a foreign language. However, the complexity and intractability of neural machine translation (NMT) models that power modern machine translation make the robustness of these systems difficult to even assess, much less guarantee. Machine translation systems can return inferior results that lead to misunderstanding, medical misdiagnoses, threats to personal safety, or political conflicts. Despite its apparent importance, validating the robustness of machine translation systems is very difficult and has, therefore, been much under-explored.
DOI 原文 ·
@inproceedings{paperbot5629,
title = {Structure-invariant testing for machine translation},
author = {Pinjia He and Clara Meister and Zhendong Su},
booktitle = {ICSE},
year = {2020},
doi = {10.1145/3377811.3380339}
}