JoelMba/MantraGSC_patents_CAS_annotations_combined_v2

该数据集是一个用于命名实体识别（NER）任务的数据集，专注于专利文本中的疾病或障碍类实体标注。数据集包含三个划分：训练集（14个示例）、验证集（4个示例）和测试集（2个示例）。每个样本包括ID、词元序列（tokens）、NER标签序列（ner_tags，使用B-Disorders、I-Disorders和O标签）以及对应的标签文本序列（ner_tag_labels）。数据来源于专利文档的标注，适用于自然语言处理中的实体识别模型训练和评估。

This dataset is designed for Named Entity Recognition (NER) tasks, focusing on the annotation of disease or disorder entities in patent texts. It includes three splits: train (14 examples), validation (4 examples), and test (2 examples). Each sample contains an ID, a sequence of tokens, a sequence of NER tags (using labels such as B-Disorders, I-Disorders, and O), and a corresponding sequence of label texts (ner_tag_labels). The data is derived from annotated patent documents and is suitable for training and evaluating NER models in natural language processing.