You Are What You Train: 数据组成对训练上下文感知机器翻译模型的影响
You Are What You Train: Effects of Data Composition on Training Context-aware Machine Translation Models
AI总结：
本文通过构建受控比例的数据集验证了训练数据稀疏性是上下文感知机器翻译的关键瓶颈，并提出两种训练策略，在单语和多语设置的 ctxPro 评估中分别提升 6 和 8 个百分点。
摘要：
实现人类水平的翻译需要利用上下文以确保连贯性，并处理代词消歧等复杂现象。标准训练数据中上下文丰富样本的稀疏性被假设为上下文利用困难的根本原因。在本研究中，我们通过构建具有受控比例的上下文相关样本的训练数据集，在单语和多语设置下系统地验证了这一假设。我们证明了训练数据稀疏性与模型性能之间存在强相关性，确认稀疏性是关键的瓶颈。重要的是，我们发现在一种上下文现象上的提升并不能泛化到其他现象。尽管我们观察到了一些跨语言迁移，但在同一语系内的语言之间，这种迁移并没有显著提高。最后，我们提出并实证评估了两种旨在利用可用数据的训练策略。这些策略改善了上下文利用，在单语和多语设置的 ctxPro 评估中分别带来了最高 6 和 8 个百分点的准确率提升。
英文摘要：
Achieving human-level translations requires leveraging context to ensure coherence and handle complex phenomena like pronoun disambiguation. Sparsity of contextually rich examples in the standard training data has been hypothesized as the reason for the difficulty of context utilization. In this work, we systematically validate this claim in both single- and multilingual settings by constructing training datasets with a controlled proportions of contextually relevant examples. We demonstrate a strong association between training data sparsity and model performance confirming sparsity as a key bottleneck. Importantly, we reveal that improvements in one contextual phenomenon do no generalize to others. While we observe some cross-lingual transfer, it is not significantly higher between languages within the same sub-family. Finally, we propose and empirically evaluate two training strategies designed to leverage the available data. These strategies improve context utilization, resulting in accuracy gains of up to 6 and 8 percentage points on the ctxPro evaluation in single- and multilingual settings respectively.
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。
