TP-NER: A named entity recognition dataset of target and precursor named entities for high-temperature steel

This dataset contains target and precursor-named entity data for high-temperature steel-related texts from published papers in IOB format. This dataset has been annotated and verified by domain experts. This dataset has 249304 rows and 3 columns. The sentence ID is in the first column, the tokens are in the second column, and the target or precursor tag is in the third column. This dataset contains data for five annotated documents among 25 documents.

本数据集汇集了源自已发表论文的高温钢相关文本中的目标实体及其前驱实体名称数据，采用IOB格式进行标注。该数据集已由领域专家进行标注与验证。数据集包含249,304行，3列。首列记录句子ID，次列列出Token，第三列标注目标或前驱实体标签。数据集覆盖了25份文档中的5份标注文档。