@inproceedings{li-zhou-2020-ji,
    title = "基于层次化语义框架的知识库属性映射方法(Property Mapping in Knowledge Base Under the Hierarchical Semantic Framework)",
    author = "Li, Yu  and
      Zhou, Guangyou",
    editor = "Sun, Maosong  and
      Li, Sujian  and
      Zhang, Yue  and
      Liu, Yang",
    booktitle = "Proceedings of the 19th Chinese National Conference on Computational Linguistics",
    month = oct,
    year = "2020",
    address = "Haikou, China",
    publisher = "Chinese Information Processing Society of China",
    url = "https://preview.aclanthology.org/jlcl-multiple-ingestion/2020.ccl-1.24/",
    pages = "246--255",
    language = "zho",
    abstract = "面向知识库的自动问答是自然语言处理的一项重要任务,它旨在对用户提出的自然语言形式问题给出精炼、准确的回复。目前由于缺少数据集、特征不一致等因素,导致难以使用通用的数据和方法实现领域知识库问答。因此,本文将{\textquotedblleft}问题意图{\textquotedblright}视作不同领域问答可能存在的共同特征,将{\textquotedblleft}问题{\textquotedblright}与三元组知识库中{\textquotedblleft}关系谓词{\textquotedblright}的映射过程作为问答核心工作。为了考虑多种层次的语义避免重要信息的损失,本文分别将{\textquotedblleft}基于门控卷积的深层语义{\textquotedblright}和{\textquotedblleft}基于交互注意力机制的浅层语义{\textquotedblright}两个方面通过门控感知机制相融合。我们在NLPCC-ICCPOL 2016 KBQA数据集上的实验表明,本文提出的方法与现有的基于CDSSM和BDSSM相比,效能有明显的提升。此外,本文通过构造天文常识知识库,将问题与关系谓词映射模型移植到特定领域,结合Bi-LSTM-CRF模型构建了天文常识自动问答系统。"
}