HuiHuang/NER-CoT
收藏Hugging Face2025-11-13 更新2025-11-15 收录
下载链接：
https://hf-mirror.com/datasets/HuiHuang/NER-CoT
下载链接
链接失效反馈官方服务：
资源简介：
ReasoningNER是一个为支持命名实体识别(NER)推理范式而设计的数据库。它包含了详细的推理轨迹（Chain-of-Thought, CoT）注释，这些注释详细说明了如何基于上下文线索、语义提示和逻辑约束来识别实体，而不是仅仅依赖语义模式。该数据集基于PILE-NER数据库构建，对其样本进行了详细的推理注释增强。
ReasoningNER is a dataset designed to support a reasoning paradigm for Named Entity Recognition (NER). It includes detailed Chain-of-Thought (CoT) annotations that explain how entities are identified based on contextual clues, semantic hints, and logical constraints, rather than relying purely on semantic patterns. This dataset is constructed based on the PILE-NER dataset, enhancing its samples with detailed reasoning annotations.
提供机构：
HuiHuang
