为多语言问答系统设计翻译质量评分机制，提升低资源语言效果
Quality-Aware Translation Tagging in Multilingual RAG system
- 引入三维度翻译质量评估，不修改原文添加评分元数据
- 在跨语言问答任务中显著优于基线模型，提升事实一致性
- 适合低资源语言场景，尤其在缺乏本地语料时实用
多语言检索增强生成（mRAG）常从英文文档检索并翻译为查询语言，用于低资源语言场景。然而，翻译质量差会降低生成效果。现有方法或假设翻译质量足够，或采用重写策略，导致事实扭曲和幻觉。为此，我们提出质量感知翻译标注（QTT-RAG），从语义等价性、语法准确性和自然流畅性三个维度显式评估翻译质量，并将评分作为元数据附加，不修改原始内容。我们在两个开放域问答基准（XORQA、MKQA）上，使用六种指令微调的大模型（参数量2.4B至14B），涵盖韩语、芬兰语（低资源）和中文（高资源），对比CrossRAG与DKM-RAG。结果表明，QTT-RAG在保持事实完整性的同时，使生成模型能基于翻译可靠性做出更优决策，有效利用跨语言文档，是低资源环境下具有普适性的可靠解决方案。
原文摘要 · Abstract (English)
Multilingual Retrieval-Augmented Generation (mRAG) often retrieves English documents and translates them into the query language for low-resource settings. However, poor translation quality degrades response generation performance. Existing approaches either assume sufficient translation quality or utilize the rewriting method, which introduces factual distortion and hallucinations. To mitigate these problems, we propose Quality-Aware Translation Tagging in mRAG (QTT-RAG), which explicitly evaluates translation quality along three dimensions-semantic equivalence, grammatical accuracy, and naturalness&fluency-and attach these scores as metadata without altering the original content. We evaluate QTT-RAG against CrossRAG and DKM-RAG as baselines in two open-domain QA benchmarks (XORQA, MKQA) using six instruction-tuned LLMs ranging from 2.4B to 14B parameters, covering two low-resource languages (Korean and Finnish) and one high-resource language (Chinese). QTT-RAG outperforms the baselines by preserving factual integrity while enabling generator models to make informed decisions based on translation reliability. This approach allows for effective usage of cross-lingual documents in low-resource settings with limited native language documents, offering a practical and robust solution across multilingual domains.
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。