? 结合触发词和论元语义的中文触发词识别三层联合模型
Journal of Computer Science and Technology
Quick Search in JCST
 Advanced Search 
      Home | PrePrint | SiteMap | Contact Us | Help
 
Indexed by   SCIE, EI ...
Bimonthly    Since 1986
Journal of Computer Science and Technology 2017, Vol. 32 Issue (5) :1044-1056    DOI: 10.1007/s11390-017-1780-5
Regular Paper << Previous Articles | >>
结合触发词和论元语义的中文触发词识别三层联合模型
Pei-Feng Li, Senior Member, CCF, Member, ACM, IEEE, Guo-Dong Zhou, Senior Member, CCF, Member, ACM, IEEE
School of Computer Science and Technology, Soochow University, Suzhou 215006, China
Three-Layer Joint Modeling of Chinese Trigger Extraction with Constraints on Trigger and Argument Semantics
Pei-Feng Li, Senior Member, CCF, Member, ACM, IEEE, Guo-Dong Zhou, Senior Member, CCF, Member, ACM, IEEE
School of Computer Science and Technology, Soochow University, Suzhou 215006, China

摘要
参考文献
相关文章
Download: [PDF 366KB]  
摘要 信息抽取的任务是从自然文本中抽取结构化信息。作为信息抽取的一个子任务,事件抽取用于识别预先定义类型的事件触发词及其论元。一般而言,事件抽取有两个子任务,分别是触发词抽取和论元抽取。当前,频繁出现的未标注触发词实例和贫内容触发词实例是中文触发词抽取的一个挑战。本文提出了一个新的三层联合模型,用于集成触发词抽取的三个部件:触发词识别、事件类型判定和事件子类型判定。该模型可以从不同负例中捕获有价值的证据,用于消除未标注触发词实例造成的负面影响。另外,本文把多种基于触发词和论元语义的约束条件引入三层联合模型,用于识别那些贫内容的事件实例。实验结果表明,我们的模型优于现有最好的中文触发词抽取和论元抽取系统。
关键词信息抽取   触发词抽取   联合模型   触发词语义   论元语义     
Abstract: As a subtask of Information Extraction (IE), which aims to extract structured information from a text, event extraction is to recognize event trigger mentions of a predefined event type and their arguments. In general, event extraction can be divided into two subtasks:trigger extraction and argument extraction. Currently, the frequent existences of un-annotated trigger mentions and poor-context trigger mentions impose critical challenges in Chinese trigger extraction. This paper proposes a novel three-layer joint model to integrate three components in trigger extraction, i.e., trigger identification, event type determination and event subtype determination. In this way, different evidence on distinct pseudo samples can be well captured to eliminate the harmful effects of those un-annotated trigger mentions. In addition, this paper introduces various types of linguistically driven constraints on trigger and argument semantics into the joint model to recover those poor-context trigger mentions. The experimental results show that our joint model significantly outperforms the state-of-the-art Chinese trigger extraction and Chinese event extraction as a whole.
Keywordsinformation extraction   trigger extraction   joint modeling   trigger semantics   argument semantics     
Received 2016-07-13;
本文基金:

The work was supported by the National Natural Science Foundation of China under Grant Nos. 61331011 and 61472265, and partially supported by Collaborative Innovation Center of Novel Software Technology and Industrialization of China.

About author: Pei-Feng Li received his Ph.D., M.S., and B.S. degrees, all in computer science from Soochow University, Suzhou, in 2006, 1997 and 1994, respectively. Currently, he is a professor at the School of Computer Science and Technology, Soochow University, Suzhou. His current research interests include Chinese information processing, machine learning and information extraction.
引用本文:   
Pei-Feng Li, Guo-Dong Zhou.结合触发词和论元语义的中文触发词识别三层联合模型[J]  Journal of Computer Science and Technology , 2017,V32(5): 1044-1056
Pei-Feng Li, Guo-Dong Zhou.Three-Layer Joint Modeling of Chinese Trigger Extraction with Constraints on Trigger and Argument Semantics[J]  Journal of Computer Science and Technology, 2017,V32(5): 1044-1056
链接本文:  
http://jcst.ict.ac.cn:8080/jcst/CN/10.1007/s11390-017-1780-5
Copyright 2010 by Journal of Computer Science and Technology