用奖励模型过滤问答中的冗余信息，提升低资源场景下的答题效率。
Context Filtering with Reward Modeling in Question Answering
- 通过奖励模型识别并保留关键信息，过滤无关内容。
- 在EPT指标上提升6.8倍，显著增强令牌效率。
- 无需人工标注，适合资源有限的QA系统优化。
自然语言处理中的问答任务需要从检索系统返回的上下文中找到答案。然而，上下文常混杂相关与不相关的信息，影响问答性能。为此，本文提出一种上下文过滤方法，利用奖励模型在摘要训练中去除非必要细节，聚焦核心内容。该方法通过分析数据对，自动区分有用信息，避免昂贵的人工评估。实验表明，该方法在EM Per Token（EPT）指标上实现6.8倍提升，我们提出此指标作为令牌效率的衡量标准，表明其在低资源场景下具有显著的效率优势。
原文摘要 · Abstract (English)
Question Answering (QA) in NLP is the task of finding answers to a query within a relevant context retrieved by a retrieval system. Yet, the mix of relevant and irrelevant information in these contexts can hinder performance enhancements in QA tasks. To address this, we introduce a context filtering approach that removes non-essential details, summarizing crucial content through Reward Modeling. This method emphasizes keeping vital data while omitting the extraneous during summarization model training. We offer a framework for developing efficient QA models by discerning useful information from dataset pairs, bypassing the need for costly human evaluation. Furthermore, we show that our approach can significantly outperform the baseline, as evidenced by a 6.8-fold increase in the EM Per Token (EPT) metric, which we propose as a measure of token efficiency, indicating a notable token-efficiency boost for low-resource settings.
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。