理解跨语言摘要中的翻译腔
Understanding Translationese in Cross-Lingual Summarization
AI总结：
本文研究跨语言摘要中翻译腔现象，确认不同数据集构建方法导致不同程度翻译腔，并系统分析其对模型评估和性能的影响，提出未来研究建议。
摘要：
给定源语言文档，跨语言摘要（CLS）旨在生成不同目标语言的简洁摘要。与单语摘要（MS）不同，自然存在的源语言文档与目标语言摘要配对的情况很少见。为了收集大规模CLS数据，现有数据集通常在创建过程中涉及翻译。然而，翻译文本与原始用该语言撰写的文本有所区别，即翻译腔。在本文中，我们首先确认构建CLS数据集的不同方法会导致不同程度的翻译腔。然后，我们系统地研究了翻译腔在源文档或目标摘要中出现时如何影响CLS模型评估和性能。具体而言，我们发现：（1）测试集中文档或摘要的翻译腔可能导致人工判断与自动评估之间的差异；（2）训练集中的翻译腔会损害模型在真实世界应用中的性能；（3）尽管机器翻译文档涉及翻译腔，但在特定训练策略下，它们对于在低资源语言上构建CLS系统非常有用。最后，我们为未来CLS研究（包括数据集和模型开发）提出建议。我们希望我们的工作能让研究人员注意到CLS中翻译腔的现象，并在未来加以考虑。
英文摘要：
Given a document in a source language, cross-lingual summarization (CLS) aims at generating a concise summary in a different target language. Unlike monolingual summarization (MS), naturally occurring source-language documents paired with target-language summaries are rare. To collect large-scale CLS data, existing datasets typically involve translation in their creation. However, the translated text is distinguished from the text originally written in that language, i.e., translationese. In this paper, we first confirm that different approaches of constructing CLS datasets will lead to different degrees of translationese. Then we systematically investigate how translationese affects CLS model evaluation and performance when it appears in source documents or target summaries. In detail, we find that (1) the translationese in documents or summaries of test sets might lead to the discrepancy between human judgment and automatic evaluation; (2) the translationese in training sets would harm model performance in real-world applications; (3) though machine-translated documents involve translationese, they are very useful for building CLS systems on low-resource languages under specific training strategies. Lastly, we give suggestions for future CLS research including dataset and model developments. We hope that our work could let researchers notice the phenomenon of translationese in CLS and take it into account in the future.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
