Page 165 - 《软件学报》2021年第12期

软件学报 ISSN 1000-9825, CODEN RUXUEW E-mail: jos@iscas.ac.cn
Journal of Software,2021,32(12):3829−3838 [doi: 10.13328/j.cnki.jos.006110] http://www.jos.org.cn
©中国科学院软件研究所版权所有. Tel: +86-10-62562563

∗
案件要素句子关联图卷积的案件舆情摘要方法

1,2
1,2
1,2
1,2
韩鹏宇 , 余正涛 , 高盛祥 , 黄于欣 , 郭军军 1,2
1
(昆明理工大学 信息工程与自动化学院,云南 昆明 650504)
2 (云南省人工智能重点实验室(昆明理工大学),云南 昆明 650504)
通讯作者: 余正涛, E-mail: ztyu@hotmail.com

摘 要: 案件舆情摘要是从涉及特定案件的新闻文本簇中,抽取能够概括其主题信息的几个句子作为摘要.案件
舆情摘要可以看作特定领域的多文档摘要,与一般的摘要任务相比,可以通过一些贯穿于整个文本簇的案件要素来
表征其主题信息.在文本簇中,由于句子与句子之间存在关联关系,案件要素与句子亦存在着不同程度的关联关系,
这些关联关系对摘要句的抽取有着重要的作用.提出了基于案件要素句子关联图卷积的案件文本摘要方法,采用图
的结构来对多文本簇进行建模,句子作为主节点,词和案件要素作为辅助节点来增强句子之间的关联关系,利用多种
特征计算不同节点间的关联关系.然后,使用图卷积神经网络学习句子关联图,并对句子进行分类得到候选摘要句.
最后,通过去重和排序得到案件舆情摘要.在收集到的案件舆情摘要数据集上进行实验,结果表明:提出的方法相比
基准模型取得了更好的效果,引入要素及句子关联图对案件多文档摘要有很好的效果.
关键词: 案件舆情摘要;图卷积;案件要素;句子关联图
中图法分类号: TP18

中文引用格式: 韩鹏宇,余正涛,高盛祥,黄于欣,郭军军.案件要素句子关联图卷积的案件舆情摘要方法.软件学报,2021,
32(12):3829−3838. http://www.jos.org.cn/1000-9825/6110.htm
英文引用格式: Han PY, Yu ZT, Gao SX, Huang YX, Guo JJ. Case-related public opinion summarization method based on graph
convolution of sentence association graph with case elements. Ruan Jian Xue Bao/Journal of Software, 2021,32(12):3829−3838
(in Chinese). http://www.jos.org.cn/1000-9825/6110.htm

Case-related Public Opinion Summarization Method Based on Graph Convolution of Sentence
Association Graph with Case Elements

1,2
1,2
1,2
1,2
HAN Peng-Yu , YU Zheng-Tao , GAO Sheng-Xiang , HUANG Yu-Xin , GUO Jun-Jun 1,2
1 (Faculty of Information Engineering and Automation, Kunming University of Science and Technology, Kunming 650504, China)
2 (Yunnan Key Laboratory of Artificial Intelligence (Kunming University of Science and Technology), Kunming 650504, China)
Abstract: The case-related public opinion summarization is the task of extracting a few sentences that can summarize the subject
information from some case-related news documents. The case-related public opinion summarization can be regarded as a multi-document
summarization in a specific field. Compared with the general multi-document summarization, the topic information can be characterized
by some case elements that run through the entire text cluster. In text clusters, sentences and sentences are associated with each other, case
elements also have associations of varying degree with sentences. These associations play an important role in extracting abstract
sentences. A case-related public opinion summarization method based on graph convolution of sentence association graph with case
elements is proposed, which uses graph structure to model all text clusters, with sentences as the main node, words and case elements as

∗ 基金项目 : 国家重点研发计 划 (2018YFC0830105, 2018YFC0830101, 2018YFC0830100); 国家自然科 学基金 (61761026,
61972186, 61762056); 云南省自然科学基金(2018FB104)
Foundation item: National Key Research and Development Project (2018YFC0830105, 2018YFC0830101, 2018YFC0830100);
National Natural Science Foundation of China (61761026, 61972186, 61762056); Natural Science Foundation of Yunnan Province
(2018FB104)
收稿时间: 2020-02-10; 修改时间: 2020-04-11, 2020-05-26; 采用时间: 2020-06-26