EBM NLP 数据集简介

该语料库包含4,993个带有(P)参与者、(I)干预和(O)结果注释的摘要。训练标签来自AMT工人，并进行了聚合以减少噪音。测试标签由医疗专业人员收集。

引用信息

@inproceedings{nye-etal-2018-corpus,
    title = "A Corpus with Multi-Level Annotations of Patients, Interventions and Outcomes to Support Language Processing for Medical Literature",
    author = "Nye, Benjamin  and
      Li, Junyi Jessy  and
      Patel, Roma  and
      Yang, Yinfei  and
      Marshall, Iain  and
      Nenkova, Ani  and
      Wallace, Byron",
    booktitle = "Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)",
    month = jul,
    year = "2018",
    address = "Melbourne, Australia",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/P18-1019",
    doi = "10.18653/v1/P18-1019",
    pages = "197--207",
}