Page 263 - 《软件学报》2021年第8期

软件学报 ISSN 1000-9825, CODEN RUXUEW E-mail: jos@iscas.ac.cn
Journal of Software,2021,32(8):2545−2556 [doi: 10.13328/j.cnki.jos.006114] http://www.jos.org.cn
©中国科学院软件研究所版权所有. Tel: +86-10-62562563

∗
基于关联记忆网络的中文细粒度命名实体识别

琚生根, 李天宁, 孙界平

(四川大学 计算机学院,四川 成都 610065)
通讯作者: 孙界平, E-mail: sunjieping@scu.edu.cn

摘 要: 细粒度命名实体识别是对文本中的实体进行定位,并将其分类至预定义的细粒度类别中.目前,中文细粒
度命名实体识别仅使用预训练语言模型对句子中的字符进行上下文编码,并没有考虑到类别的标签信息具有区分
实体类别的能力.由于预测句子不带有实体标签,使用关联记忆网络来捕获训练集句子的实体标签信息,并将标签信
息融入预测句子的字符表示中.该方法将训练集中带实体标签的句子作为记忆单元,利用预训练语言模型获取原句
子和记忆单元句子的上下文表示,再通过注意力机制将记忆单元句子的标签信息与原句子的表示结合,从而提升识
别效果.在 CLUENER 2020 中文细粒度命名实体识别任务上,该方法对比基线方法获得了提升.
关键词: 中文细粒度命名实体识别;关联记忆网络;多头自注意力;预训练语言模型
中图法分类号: TP18

中文引用格式: 琚生根,李天宁,孙界平.基于关联记忆网络的中文细粒度命名实体识别.软件学报,2021,32(8):2545−2556.
http://www.jos.org.cn/1000-9825/6114.htm
英文引用格式: Ju SG, Li TN, Sun JP. Chinese fine-grained name entity recognition based on associated memory networks. Ruan
Jian Xue Bao/Journal of Software, 2021,32(8):2545−2556 (in Chinese). http://www.jos.org.cn/1000-9825/6114.htm
Chinese Fine-grained Name Entity Recognition Based on Associated Memory Networks

JU Sheng-Gen, LI Tian-Ning, SUN Jie-Ping
(College of Computer Science, Sichuan University, Chengdu 610065, China)

Abstract: Fine-grained named entity recognition is to locate entities in text and classify them into predefined fine-grained categories. At
present, Chinese fine-grained named entity recognition only uses pre-trained language models to encode characters in sentences and does
not take into account that the category label information can distinguish entity categories. Since the predicted sentence does not have the
entity label, the associated memory network is used to capture the entity label information of the sentences in the training set and to
incorporate label information into the representation of predicted sentences in this paper. In this method, sentences with entity labels in
the training set are used as memory units, the pre-trained language model is used to obtain the contextual representations of the original
sentence and the sentence in the memory unit. Then, the label information of the sentences in the memory unit is combined with the
representation of the original sentence by the attention mechanism to improve the recognition effect. On the CLUENER 2020 Chinese
fine-grained named entity recognition task, this method improves performance over the baseline methods.
Key words: Chinese fine-grained name entity recognition; associated memory network; multi-headed self-attention; pre-trained
language model

命名实体识别是自然语言处理中的信息抽取任务之一,其目的是对文本中特定类别的实体进行定位和分

∗ 基金项目: 国家自然科学基金(61972270); 四川省新一代人工智能重大专项(2018GZDZX0039); 四川省重点研发项目(2019
YFG0521)
Foundation item: National Natural Science Foundation of China (61972270); New Generation of Artificial Intelligence Major
Project of Sichuan Province (2018GZDZX0039); Major Science and Technology Research and Development Plan of Sichuan Province
(2019YFG0521)
收稿时间: 2020-04-13; 修改时间: 2020-05-27, 2020-06-20; 采用时间: 2020-07-01; jos 在线出版时间: 2021-04-20