arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2410.07779cs.CL
收藏

使用自动指标建模用户偏好:为机器翻译创建高质量偏好数据集

Modeling User Preferences with Automatic Metrics: Creating a High-Quality Preference Dataset for Machine Translation

Instituto de Telecomunicações(电信研究所);Unbabel(Unbabel公司);Instituto Superior Técnico, Universidade de Lisboa(里斯本大学高等技术学院);Carnegie Mellon University(卡内基梅隆大学);CentraleSupélec, Université Paris-Saclay(巴黎萨克雷大学中央高等电力学院);ELLIS Unit Lisbon(里斯本ELLIS单元)

Sweta Agrawal, José G. C. de Souza, Ricardo Rei, António Farinhas, Gonçalo Faria, Patrick Fernandes, Nuno M Guerreiro, Andre Martins

AI总结:

本文提出结合人类专家评估与自动指标的方法,构建了覆盖18个语言方向的MT-Pref数据集,并在WMT23和FLORES基准上验证其对齐TOWER模型后显著提升机器翻译质量。

摘要:

与人类偏好对齐是开发准确且安全的大型语言模型的重要步骤。在机器翻译(MT)中也不例外,更好地处理语言细微差别和特定语境的变化能提升翻译质量。然而,基于人类反馈的偏好数据在大规模获取和整理时成本极高。另一方面,自动指标虽然能够诱导出偏好,但可能无法完全匹配人类期望。在本文中,我们提出了一种结合两者优势的方法。我们首先收集专业语言学家对多个高质量MT系统生成翻译的句子级质量评估,并评估当前自动指标恢复这些偏好的能力。随后,我们利用这一分析整理出一个新数据集——MT-Pref(metric induced translation preference,指标诱导翻译偏好)数据集,该数据集包含18k个实例,覆盖18个语言方向,使用2022年之后的多领域文本。我们证明,在MT-Pref上对齐TOWER模型能显著提升WMT23和FLORES基准上的翻译质量。

英文摘要:

Alignment with human preferences is an important step in developing accurate and safe large language models. This is no exception in machine translation (MT), where better handling of language nuances and context-specific variations leads to improved quality. However, preference data based on human feedback can be very expensive to obtain and curate at a large scale. Automatic metrics, on the other hand, can induce preferences, but they might not match human expectations perfectly. In this paper, we propose an approach that leverages the best of both worlds. We first collect sentence-level quality assessments from professional linguists on translations generated by multiple high-quality MT systems and evaluate the ability of current automatic metrics to recover these preferences. We then use this analysis to curate a new dataset, MT-Pref (metric induced translation preference) dataset, which comprises 18k instances covering 18 language directions, using texts sourced from multiple domains post-2022. We show that aligning TOWER models on MT-Pref significantly improves translation quality on WMT23 and FLORES benchmarks.

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。

↑