事理图谱是一种描述事件之间顺承、因果等关系的事理演化逻辑有向图,它蕴含了丰富的事件间关系,在各领域都具有重要的研究意义和应用价值。当前研究主要集中于公开域的事件抽取上,而在特定领域,如政治领域,因其事件类型和事件内容较为复杂,相关研究十分有限。该文旨在构建面向政治领域的事理图谱,针对政治事件抽取中存在的语料匮乏、标准缺失等问题,制定了一套面向政治领域的事件分类标准,构建了一套政治领域的事件语料库。同时,该文分别提出了一种融合注意力机制的字嵌入修正神经网络的Pipeline模型和一种基于BERT+BiLSTM的Joint模型进行事件触发词和论元抽取,并在该语料库上进行实验。实验结果表明,两种模型在事件触发词与论元抽取任务中,F1指标较基线模型均有较大提升。
Abstract
The Eventic Graph(EG) is a directed graph that describes the logic relationship between events, such as continuation and causality, etc. In contrast to the current researches focusing on the event extraction on the open domain. this paper aims at the construction of the eventic graph for the political field. We establish an annotation scheme for political events and construct an event corpus for the political field. Moreover, we present a character embedding based neural network by integrating the attention mechanism and a BERT+BiLSTM framework for political event extraction as the pipeline and the joint model, respectively. Experiments on our constructed corpus show that the porposed method could achieve significant improvement on event classification and argument classification in terms of F1-score compared with previous neural network based methods.
关键词
事理图谱 /
字嵌入修正神经网络 /
事件抽取
{{custom_keyword}} /
Key words
eventic graph /
character embedding based neural network /
event extraction
{{custom_keyword}} /
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] 李忠阳,赵森栋,丁效.事理图谱:事件演化的规律和模式[EB/OL].https://www.sohu.com/a/140664312_657157[2017-05-15].
[2] 欧阳康,张名仓.社会科学研究方法[M].北京:高等教育出版社,2001.
[3] Linguistic Data Consortium, ACE (Automatic Content Extraction) Chinese Annotation Guideline for Event[EB/OL], LDC,2005.https://www.doc88.com/p-180734447445.html
[4] Chen Y, Xu L, Liu K, et al. Event extraction via dynamic multi-pooling convolutional neural net-works[C]//Proceedings of ACL, 2015:167-176.
[5] Lin H, Lu Y, Han X, et al. Nugget proposal networks for Chinese event detection[C]//Proceedings of ACL, 2018:1565-1574.
[6] Jacob Devlin, Mingwei Chang, Kenton Lee, et al. BERT: Pre-training of deep bidirectional transformers for language understanding[J].arXiv:1810.04805v2[cs.CL],2019.
[7] 白璐.面向政治领域的事理演化图谱构建[D].北京:国际关系学院硕士学位论文,2020.
[8] Yang C C, Shi X. Discovering event evolution graphs from newswires[C]//Proceedings of the 15th International Conference on World Wide Web, 2006: 945-946.
[9] 周京艳,刘如,李佳娱,等.情报事理图谱的概念界定与价值分析[J].情报杂志,2018,37(05):31-36.
[10] Hogenboom F, Frasincar F, Kaymak U. An overview of event extraction from text[J]. CEUR-WS. Org, 2011, 779: 48-57.
[11] Yakushiji A, Tateisi Y, Miyao Y. Event extraction from biomedical papers using a full parser[J]. Pacific Symposium on Biocomputing, 2001, 6:408-419.
[12] Li P, Zhou G. Employing Morphological structures and sememes for Chinese event extraction[C]//Proceedings of COLING,2012.
[13] Zhao K. Structured prediction with perceptron: Theory and algorithms[EB/OL].https://www.gc.cuny.edu/CUNY_GC/media/Computer-Science/Student%20Presentations/Kai%20Zhao/Second_Exam_Survey_Kai_Zhao_12_11_2014.pdf[2014-12-11].
[14] Nguyen T H, Grishman R. Event detection and domain adaptation with convolutional neural networks[C]//Proceedings of the 53rd ACL, 2015.
[15] 杨健.基于事件抽取的日汉新闻语料库构建研究[D].北京:北京交通大学硕士学位论文,2016.
[16] 干红华,潘云鹤.一种基于事件的因果关系的结构分析方法[J].模式识别与人工智能, 2003, 16(01):56-62.
[17] Marcu D, Echihabi A. An unsupervised approach to recognizing discourse relations[C]//Proceedings of the 40th ACL, 2002:368-375.
[18] Sorgente A, Vettigli G, Mele F. Automatic extraction of cause-effect relations in natural language text[J].CEUR-WS, 2013:37-48.
[19] 吴春子.面向互联网新闻的重点人物识别及相关事件演化的研究与实现[D].北京:北京邮电大学硕士学位论文, 2018.
[20] 杨竣辉,刘宗田,刘炜,等.基于语义事件因果关系识别[J].小型微型计算机系统,2016,37(3):433-437.
[21] 张晓斌,陈福才,黄瑞阳.基于CNN和双向LSTM融合的实体关系抽取[J].网络与信息安全学报,2018,4(09): 44-51.
[22] Chen C, Ng V. Joint modeling for Chinese event extraction with rich linguistic features[C]//Proceedings of the COLING,2012:529-544.
[23] Chambers N, Jurafsky D. Unsupervised Learning of Narrative Event Chains[C]//Proceedings of the ACL,2008:789-797.
[24] Feng X, Qin B, Liu T. A language-independent neural network for event detection[J]. Science China In-formation Sciences, 2018, 61(9): 92-106.
[25] Ghaeini R, Fern X Z, Huang L. Event nugget detection with forward-backward recurrent neural networks[J/OL]. arXiv preprint arXiv:1802.05672, 2018.
[26] Li L, Wan J, Zheng J. Biomedical event extraction based on GRU integrating attention mechanism[J]. BMC bioinformatics, 2018, 19(9): 177.
[27] Li P, Zhu Q, Zhou G. Argument inference from relevant event mentions in Chinese argument extraction[C]//Proceedings of the ACL,2013.
[28] Li Z, Ding X, Liu T. Constructing narrative event evolutionary graph for script event prediction[C]//Proceedings of International Joint Conferences on Artificial Intelligence Organization,2018.
[29] Nguyen T M, Nguyen T H. One for All: Neural Joint Modeling of Entities and Events[C]//Proceedings of the AAAI, 2019.
[30] Pichotta K, Mooney R J. Using sentence-level LSTM language models for script inference[C]//Proceedings of the ACL, 2016.
[31] Sha L, Qian F, Chang B. Jointly extracting event triggers and arguments by dependency-bridge RNN and tensor-based argument interaction[C]//Proceedings of the 32nd AAAI Conference on Artificial Intelligence, 2018.
[32] Verga P, Strubell E, McCallum A. Simultaneously self-attending to all mentions for full-abstract biological relation extraction[J/OL]. arXiv preprint arXiv:1802.10569, 2018.
[33] Yakushiji A , Tateisi Y , Miyao Y. Event extraction from biomedical papers using a full parser.[J]. Pacific Symposium on Biocomputing. Pacific Symposium on Biocomputing, 2001, 6:408-419.
[34] Liu X. Jointly multiple events extraction via attention-based Graph Information Aggregation[C]//Proceedings of the EMNLP, 2018.
[35] Yang S. Exploring Pre-trained language models for event extraction and generation[C]//Proceedings of the ACL, 2019.
{{custom_fnGroup.title_cn}}
脚注
{{custom_fn.content}}
基金
国家自然科学基金(61976066);北京市自然科学基金(4212031);中央高校基础研究课题(3262021T23);北京市科研计划课题(Z191100004019007)
{{custom_fund}}
PDF(5967 KB)
