基于注意力机制的双向长短时记忆网络模型突发事件演化关系抽取

doi:10.11772/j.issn.1001-9081.2018122533

计算机应用 ›› 2019, Vol. 39 ›› Issue (6): 1646-1651.DOI: 10.11772/j.issn.1001-9081.2018122533

基于注意力机制的双向长短时记忆网络模型突发事件演化关系抽取

闻畅^1,2,3,4, 刘宇^1,2,3,4, 顾进广^1,2,3,4

1. 武汉科技大学计算机科学与技术学院, 武汉 430065;
2. 智能信息处理与实时工业系统湖北省重点实验室(武汉科技大学), 武汉 430065;
3. 武汉科技大学大数据科学与工程研究院, 武汉 430065;
4. 国家新闻出版署富媒体数字出版内容组织与知识服务重点实验室, 北京 100038

收稿日期:2018-12-24 修回日期:2019-03-10 发布日期:2019-06-17 出版日期:2019-06-10
通讯作者: 刘宇
作者简介:闻畅(1994-),女,湖北武汉人,硕士研究生,主要研究方向:知识图谱、智能信息处理、语义Web;刘宇(1980-),男,湖北武汉人,副教授,博士,主要研究方向:知识工程、智能系统、分布式计算;顾进广(1974-),男,湖北武汉人,教授,博士,CCF会员,主要研究方向:语义Web、新型网络计算、智能信息处理。
基金资助:
国家自然科学基金资助项目（61673304）；国家社会科学基金重大计划项目（11&ZD189）。

Evolution relationship extraction of emergency based on attention-based bidirectional long short-term memory network model

WEN Chang^1,2,3,4, LIU Yu^1,2,3,4, GU Jinguang^1,2,3,4

1. College of Computer Science and Technology, Wuhan University of Science and Technology, Wuhan Hubei 430065, China;
2. Key Laboratory of Intelligent Information Processing and Real-time Industrial System in Hubei Province(Wuhan University of Science and Technology), Wuhan Hubei 430065, China;
3. Institute of Big Data Science and Engineering, Wuhan University of Science and Technology, Wuhan Hubei 430065, China;
4. Key Laboratory of Rich-media Knowledge Organization and Service of Digital Publishing Content, National Press and Publication Administration, Beijing 100038, China

Received:2018-12-24 Revised:2019-03-10 Online:2019-06-17 Published:2019-06-10
Supported by:
This work is partially supported by the National Natural Science Foundation of China (61673304), the National Social Science Fund Major Plan of China (11&ZD189).

摘要/Abstract

摘要： 针对现有突发事件关系抽取研究多集中于因果关系抽取而忽略了其他演化关系的问题，为了提高应急决策中信息抽取的完备性，应用一种基于注意力机制的双向长短时记忆（LSTM）网络模型进行突发事件演化关系抽取。首先，结合突发事件演化关系的概念，构建演化关系模型并进行形式化定义，依据模型对突发事件语料进行标注；其次，搭建双向LSTM网络结构，并引入注意力机制计算注意力概率以突出关键词汇在文本中的重要程度；最终，使用搭建的网络模型进行演化关系抽取得到结果。在演化关系抽取实验中，相对于现有因果关系抽取方法，所提方法不仅抽取出更加充分的演化关系，为突发事件应急决策提供了更完善的信息；同时，在正确率、召回率和F1分数上分别平均提升了7.3%、6.7%和7.0%，有效提高了突发事件演化关系抽取的准确性。

关键词: 关系抽取, 突发事件, 演化关系, 注意力机制, 双向长短时记忆网络

Abstract: Concerning the problem that existing study of emergency relationship extraction mostly focuses on causality extraction while neglects other evolutions, in order to improve the completeness of information extracted in emergency decision-making, a method based on attention-based bidirectional Long Short-Term Memory (LSTM) model was used to extract the evolution relationship. Firstly, combined with the concept of evolution relationship in emergencies, an evolution relationship model was constructed and given the formal definition, and the emergency corpus was labeled according to the model. Then, a bidirectional LSTM network was built and attention mechanism was introduced to calculate the attention probability to highlight the importance of the key words in the text. Finally, the built network model was used to extract the evolution relationship. In the evolution relationship extraction experiments, compared with the existing causality extraction methods, the proposed method can extract more sufficient evolution relationship for emergency decision-making. At the same time, the average precision, recall and F1_score are respectively increased by 7.3%, 6.7% and 7.0%, which effectively improves the accuracy of the evolution relationship extraction of emergency.

Key words: relationship extraction, emergency, evolutionary relation, attention mechanism, bidirectional Long Short-Term Memory (LSTM)

中图分类号:

TP182

闻畅, 刘宇, 顾进广. 基于注意力机制的双向长短时记忆网络模型突发事件演化关系抽取[J]. 计算机应用, 2019, 39(6): 1646-1651.

WEN Chang, LIU Yu, GU Jinguang. Evolution relationship extraction of emergency based on attention-based bidirectional long short-term memory network model[J]. Journal of Computer Applications, 2019, 39(6): 1646-1651.

参考文献

[1] 李湖生.安全与应急管理学科领域的概念视图及主要研究内容[J].安全,2017,38(10):23-26.(LI H S. Conceptual view and main research contents in the field of safety and emergency management[J]. Security, 2017, 38(10):23-26.)
[2] 陈玉芳,屠兢,任冬林.高校社会安全类突发事件情景要素识别提取[J].电子科技大学学报(社科版),2017,19(6):56-59.(CHEN Y F, TU J, REN D L. Identification and extraction of scene factors for social security emergencies in colleges and universities[J]. Journal of University of Electronic Science and Technology of China (Social Sciences Edition), 2017, 19(6):56-59.)
[3] KAPLAN R M, BERRY-ROGGHE G. Knowledge-based acquisition of causal relationships in text[J]. Knowledge Acquisition, 1991, 3(3):317-337.
[4] KHOO C S G, CHAN S, NIU Y. Extracting causal knowledge from a medical database using graphical patterns[C]//Proceedings of the 38th Annual Meeting for Computational Linguistics. Stroudsburg, PA:Association for Computational Linguistics, 2000:336-343.
[5] 李悦群,毛文吉,王飞跃.面向领域开源文本的因果知识提取[J].计算机工程与科学,2010,32(5):100-104.(LI Y Q, MAO W J, WANG F Y. Causal knowledge extraction based on open source domain texts[J]. Computer Engineering and Science, 2010, 32(5):100-104)
[6] GIRJU R. Automatic detection of causal relations for question answering[C]//Proceedings of the ACL 2003 Workshop on Multilingual Summarization and Question Answering. Stroudsburg, PA:Association for Computational Linguistics, 2003:76-83.
[7] 付剑锋,刘宗田,刘炜,等.基于层叠条件随机场的事件因果关系抽取[J].模式识别与人工智能,2011,24(4):567-573.(FU J F, LIU Z T, LIU W, et al. Event causality extraction based on cascaded conditional random fields[J]. Pattern Recognition & Artificial Intelligence, 2011, 24(4):567-573.)
[8] 钟军,禹龙,田生伟,等.基于双层模型的维吾尔语突发事件因果关系抽取[J].自动化学报,2014,40(4):771-779.(ZHONG J, YU L, TIAN S W, et al. Causal relationship extraction of Uyghur emergency events based on cascaded model[J]. Acta Automatica Sinica, 2014, 40(4):771-779.)
[9] 田生伟,周兴发,禹龙,等.基于双向LSTM的维吾尔语事件因果关系抽取[J].电子与信息学报,2018,40(1):200-208.(TIAN S W, ZHOU X F, YU L, et al. Causal relation extraction of Uighur events based on bidirectional long short-term memory model[J]. Journal of Electronics & Information Technology, 2018, 40(1):200-208.)
[10] BAHDANAU D, CHO K, BENGIO Y. Neural machine translation by jointly learning to align and translate[EB/OL].[2018-10-12]. https://arxiv.org/pdf/1409.0473.pdf.
[11] 盛勇,孙庆云,王永明.突发事件情景演化及关键要素提取方法[J].中国安全生产科学技术,2015,11(1):17-21.(SHENG Y, SUN Q Y, WANG Y M. Emergency scenario evolution extraction of key elements[J]. Journal of Safety Science and Technology, 2015, 11(1):17-21.)
[12] 罗鹏志,陈曦.面向事件链的突发事件建模与检索技术[J].华中科技大学学报(自然科学版),2015(S1):440-444.(LUO P Z, CHEN X. Research on emergency event chain modeling and retrieval[J]. Journal of Huazhong University of Science and Technology (Natural Science), 2015(S1):440-444.)
[13] 马建华,陈安.突发事件的演化模式分析[J].安全,2009,30(12):1-4.(MA J H, CHEN A. Analysis of the evolution model of emergency events[J]. Security, 2009, 30(12):1-4.)
[14] COMFORT L K, WANGH W L Jr, CIGLER B A. Emergency management research and practice in public administration:emergence,evolution,expansion, and future directions[J]. Public Administration Review, 2012, 72(4):539-547.
[15] OSEI-AMPOFO M, ODURO G, OTENG R, et al. The evolution and current state of emergency care in Ghana[J]. African Journal of Emergency Medicine, 2013, 3(2):52-58.
[16] MENTZELOPOULOS S D, MANTZANAS M, BELLE G, et al. Evolution of European union legislation on emergency research[J]. Resuscitation, 2015, 91:84-91.
[17] WEN C, LIU Y, GU J, et al. E-SKB:a semantic knowledge base for emergency[C]//Proceedings of the 2016 China Conference on Knowledge Graph & Semantic Computing, CCIS 650. Berlin:Springer, 2016:196-202.
[18] CAI R, ZHANG X, WANG H. Bidirectional recurrent convolutional neural network for relation classification[C]//Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics. Stroudsburg, PA:Association for Computational Linguistics, 2016:756-765.
[19] MIWA M, BANSAL M. End-to-end relation extraction using LSTMs on sequences and tree structures[C]//Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics. Stroudsburg, PA:Association for Computational Linguistics, 2016:1105-1116.
[20] ZHANG S, ZHENG D, HU X, et al. Bidirectional long short-term memory networks for relation classification[C]//Proceedings of the 29th Pacific Asia Conference on Language, Information and Computation. Stroudsburg, PA:Association for Computational Linguistics, 2015:73-78.
[21] 刘伟,陈鸿昶,黄瑞阳.基于Tree-based CNN的关系抽取[J].中文信息学报,2018,32(11):34-40.(LIU W, CHEN H C, HUANG R Y. Relation extraction via tree-based CNN[J]. Journal of Chinese Information Processing, 2018, 32(11):34-40.)

基于注意力机制的双向长短时记忆网络模型突发事件演化关系抽取

Evolution relationship extraction of emergency based on attention-based bidirectional long short-term memory network model

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

[1]	秦璟, 秦志光, 李发礼, 彭悦恒. 基于概率稀疏自注意力神经网络的重性抑郁疾患诊断[J]. 《计算机应用》唯一官方网站, 2024, 44(9): 2970-2974.
[2]	李力铤, 华蓓, 贺若舟, 徐况. 基于解耦注意力机制的多变量时序预测模型[J]. 《计算机应用》唯一官方网站, 2024, 44(9): 2732-2738.
[3]	赵志强, 马培红, 黑新宏. 基于双重注意力机制的人群计数方法[J]. 《计算机应用》唯一官方网站, 2024, 44(9): 2886-2892.
[4]	薛凯鹏, 徐涛, 廖春节. 融合自监督和多层交叉注意力的多模态情感分析网络[J]. 《计算机应用》唯一官方网站, 2024, 44(8): 2387-2392.
[5]	汪雨晴, 朱广丽, 段文杰, 李书羽, 周若彤. 基于交互注意力机制的心理咨询文本情感分类模型[J]. 《计算机应用》唯一官方网站, 2024, 44(8): 2393-2399.
[6]	高鹏淇, 黄鹤鸣, 樊永红. 融合坐标与多头注意力机制的交互语音情感识别[J]. 《计算机应用》唯一官方网站, 2024, 44(8): 2400-2406.
[7]	李钟华, 白云起, 王雪津, 黄雷雷, 林初俊, 廖诗宇. 基于图像增强的低照度人脸检测[J]. 《计算机应用》唯一官方网站, 2024, 44(8): 2588-2594.
[8]	莫尚斌, 王文君, 董凌, 高盛祥, 余正涛. 基于多路信息聚合协同解码的单通道语音增强[J]. 《计算机应用》唯一官方网站, 2024, 44(8): 2611-2617.
[9]	赵宇博, 张丽萍, 闫盛, 侯敏, 高茂. 基于改进分段卷积神经网络和知识蒸馏的学科知识实体间关系抽取[J]. 《计算机应用》唯一官方网站, 2024, 44(8): 2421-2429.
[10]	唐媛, 陈艳平, 扈应, 黄瑞章, 秦永彬. 基于多尺度混合注意力卷积神经网络的关系抽取模型[J]. 《计算机应用》唯一官方网站, 2024, 44(7): 2011-2017.
[11]	熊武, 曹从军, 宋雪芳, 邵云龙, 王旭升. 基于多尺度混合域注意力机制的笔迹鉴别方法[J]. 《计算机应用》唯一官方网站, 2024, 44(7): 2225-2232.
[12]	李欢欢, 黄添强, 丁雪梅, 罗海峰, 黄丽清. 基于多尺度时空图卷积网络的交通出行需求预测[J]. 《计算机应用》唯一官方网站, 2024, 44(7): 2065-2072.
[13]	毛典辉, 李学博, 刘峻岭, 张登辉, 颜文婧. 基于并行异构图和序列注意力机制的中文实体关系抽取模型[J]. 《计算机应用》唯一官方网站, 2024, 44(7): 2018-2025.
[14]	刘丽, 侯海金, 王安红, 张涛. 基于多尺度注意力的生成式信息隐藏算法[J]. 《计算机应用》唯一官方网站, 2024, 44(7): 2102-2109.
[15]	徐松, 张文博, 王一帆. 基于时空信息的轻量视频显著性目标检测网络[J]. 《计算机应用》唯一官方网站, 2024, 44(7): 2192-2199.