将翻译质量评估融入解码过程，提升译文整体质量。
Quality-Aware Decoding: Unifying Quality Estimation and Decoding
- 构建逐标记质量评估模型，实时评分部分译文。
- 解码时融合质量评分，比现有重排序方法提升1.39分XCOMET-XXL。
- 适合文档级翻译，尤其改善N-best列表质量不佳问题。
神经机器翻译中的质量评估（QE）模型可在无参考译文情况下预测输出质量。近年来，利用QE模型实现质量感知解码成为研究热点，其与人工判断高度相关，并能提升翻译效果。尽管已有方法通过采样多个候选译文并选取最优者来改进结果，但尚未将QE模型直接嵌入解码流程。本文提出一种新型逐标记质量评估模型，可对部分译文进行可靠评分。该模型采用单向结构，适配解码器在部分序列上的训练与高效性。进一步设计了一种融合该模型的质量感知解码策略，实验表明，相比使用最先进QE模型的N-best重排序方法，翻译质量最高提升1.39分（XCOMET-XXL↑）。此外，在文档翻译任务中，本方法显著优于传统N-best列表，因其在长文本中更易出现质量不均问题。代码见：https://ai4lt.iar.kit.edu/english/projects_kontextmt.php
原文摘要 · Abstract (English)
Quality Estimation (QE) models for Neural Machine Translation (NMT) predict the quality of the hypothesis without having access to the reference. An emerging research direction in NMT involves the use of QE models, which have demonstrated high correlations with human judgment and can enhance translations through Quality-Aware Decoding. Although several approaches have been proposed based on sampling multiple candidate translations and picking the best candidate, none have integrated these models directly into the decoding process. In this paper, we address this by proposing a novel token-level QE model capable of reliably scoring partial translations. We build a uni-directional QE model for this, as decoder models are inherently trained and efficient on partial sequences. We then present a decoding strategy that integrates the QE model for Quality-Aware decoding and demonstrate that the translation quality improves when compared to the N-best list re-ranking with state-of-the-art QE models (up to $1.39$ XCOMET-XXL $\uparrow$). Finally, we show that our approach provides significant benefits in document translation tasks, where the quality of N-best lists is typically suboptimal. Code can be found at https://ai4lt.iar.kit.edu/english/projects\_kontextmt.php
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。