不用标注数据，用多个检索器自动修正虚假信息。
Mask-to-Correct$^+$: Leveraging Retriever Diversity for Masking-guided Faithful Fact Correction
- 用多样检索器生成掩码，定位错误内容
- 结合多检索器结果，纠正后忠实度提升14%
- 无需人工标注，适合跨领域事实纠错
社交媒体上虚假信息快速传播，亟需鲁棒的自动化事实纠错框架。现有方法依赖人工标注的论点-证据对，数据稀缺且易有偏见，限制了跨领域泛化能力。同时，其修正过程常忽视语义忠实性。为此，我们提出M₂C，一种无需训练、仅在推理阶段运行的检索增强生成框架，通过感知多样性的掩码机制识别论点中的错误片段，并利用检索到的证据评估修正的忠实性。由于RAG效果高度依赖检索器选择，我们进一步提出M₂C⁺，采用多检索器集成策略，融合不同排序器的修正结果，降低检索偏差，提升鲁棒性。在基准数据集上的大量实验表明，所提框架持续优于所有基线，最高在SARI分数上提升14%，且不依赖真实证据。
原文摘要 · Abstract (English)
The rapid spread of misinformation on social media highlights the need for robust, automated fact correction frameworks. However, existing works rely on supervised learning from manually annotated claim-evidence pairs, which are scarce and prone to biases, limiting their generalization across domains. Moreover, these methods overlook semantic faithfulness in their correction process. To address these challenges, we propose Mask-to-Correct (M$_2$C), a training-free, inference-only Retrieval Augmented Generation (RAG) based framework that leverages diversity-aware masking to identify erroneous spans of claims and evaluate the faithfulness of corrections using retrieved evidence. However, the effectiveness of RAG heavily depends on the choice of retriever, which may vary across queries. To mitigate this, we further introduce M$_2$C$^+$, an ensemble-based framework that combines corrections across multiple rankers to reduce retrieval bias and improve robustness. Extensive experiments on the benchmark datasets demonstrate that our proposed frameworks consistently outperform all baselines, achieving up to 14% improvement in SARI scores, without using gold evidence.
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。