Published Versions 1 Vol 3 (3) : 444-459 2021
Download
A Joint Learning Framework for the CCKS-2020 Financial Event Extraction Task
: 2021 - 02 - 13
: 2021 - 04 - 22
: 2021 - 04 - 30
123 3 0
Abstract & Keywords
Abstract: This paper presents a winning solution for the CCKS-2020 financial event extraction task, where the goal is to identify event types, triggers and arguments in sentences across multiple event types. In this task, we focus on resolving two challenging problems (i.e., low resources and element overlapping) by proposing a joint learning framework, named SaltyFishes. We first formulate the event extraction task as a joint probability model. By sharing parameters in the model across different types, we can learn to adapt to low-resource events based on high-resource events. We further address the element overlapping problems by a mechanism of Conditional Layer Normalization, achieving even better extraction accuracy. The overall approach achieves an F1-score of 87.8% which ranks the first place in the competition.
Keywords: Event detection; Event extraction; Joint learning; Financial event.
[1]
Chen, Y., et al.: Event extraction via dynamic multi-pooling convolutional neural networks. In: Proceedings of ACL, pp. 167–176 (2015)
[2]
Deng, S., et al.: Meta-learning with dynamic-memory-based prototypical network for few-shot event detection. In: Proceedings of WSDM, pp. 151–159 (2020)
[3]
Huang, L., et al.: Zero-shot transfer learning for event extraction. In: Proceedings of ACL, pp. 2160–2170 (2018)
[4]
Nguyen, T.H., Cho, K., Grishman, R.: Joint event extraction via recurrent neural networks. In: Proceedings of NAACL, pp. 300–309 (2016)
[5]
Nguyen, T.M., Nguyen, T.H.: One for all: Neural joint modeling of entities and events. In: Proceedings of AAAI, pp. 6851–6858 (2019)
[6]
Devlin, J., et al.: BERT: pre-training of deep bidirectional transformers for language understanding. In: Proceedings of NAACL-HLT, pp. 4171–4186 (2019)
[7]
Li, Q., Ji, H., Huang, L.: Joint event extraction via structured prediction with global features. In: Proceedings of ACL. pp. 73–82 (2013)
[8]
Liu, X., Luo, Z., Huang, H.: Jointly multiple events extraction via attention-based graph information aggregation. In: Proceedings of EMNLP, pp. 1247–1256 (2018)
[9]
Sha, L., et al.: Jointly extracting event triggers and arguments by dependency-bridge RNN and tensor-based argument interaction. In: Proceedings of AAAI, pp. 5916–5923 (2018)
[10]
Du, X., Cardie, C.: Event extraction by answering (almost) natural questions. In: Proceedings of EMNLP, pp. 671–683 (2020)
[11]
Li, F., et al.: Event extraction as multi-turn question answering. In: Proceedings of EMNLP, pp. 829–838 (2020)
[12]
Liu, J., et al.: Event extraction as machine reading comprehension. In: Proceedings of EMNLP, pp. 1641–1651 (2020)
[13]
Wadden, D., et al.: Entity, relation, and event extraction with contextualized span representations. In: Proceedings of EMNLP,pp. 5783–5788 (2019)
[14]
Yang, S., et al.: Exploring pre-trained language models for event extraction and generation. In: Proceedings of ACL, pp. 5284–5294 (2019)
[15]
Cao, P., et al.: Incremental event detection via knowledge consolidation networks. In: Proceedings of EMNLP, pp. 707–717 (2020)
[16]
Su, J.L.: Conditional text generation based on conditional layer normalization (in Chinese). Availablet at: https://spaces.ac.cn/archives/7124. Accessed 21 May 2021
[17]
Yu, B., et al.: Semi-open information extraction. In: Proceedings of WWW (2021). Available at: https://www2021.thewebconf.org/program/papers/. Accessed 21 May 2021
[18]
Ba, L.J., Kiros, J.R., Hinton, G.E.: Layer normalization. arXiv preprint arXiv:1607.06450 (2016)
Article and author information
Cite As
Citation: Sheng, J.W., et al: A joint learning framework for the CCKS-2020 financial event extraction task. Data Intelligence 3(3), 444-459 (2021). doi: 10.1162/dint_a_00098
Jiawei Sheng
This work was a collaboration of all the authors. J.W. Sheng proposed the idea of the joint learning framework, and was the team leader in the CCKS-2020 competition. All the authors have made meaningful and valuable contributions to this manuscript.
Jiawei Sheng is currently pursuing his PhD degree in the Institute ofInformation Engineering, Chinese Academy of Sciences. His current researchinterests include information extraction, knowledge graph embedding andknowledge acquisition.
0000-0002-4865-982X
Qian Li
Q. Li and Y.M. Hei devised and implemented the details of the model. All the authors have made meaningful and valuable contributions to this manuscript.
Qian Li is currently pursuing her PhD degree in the School of ComputerScience, Beihang University. Her current research interests include knowledgegraph embedding and information extraction.
0000-0002-1612-4644
Yiming Hei
Q. Li and Y.M. Hei devised and implemented the details of the model. All the authors have made meaningful and valuable contributions to this manuscript.
Yiming Hei is currently pursuing his PhD degree in the School of CyberScience and Technology, Beihang University. His research interests includeknowledge graph embedding and information extraction.
0000-0003-0794-9932
Shu Guo
S. Guo revised and proofread the resulting manuscript. All the authors have made meaningful and valuable contributions to this manuscript.
guoshu@cert.org.cn
Shu Guo received her PhD degree from the Institute of InformationEngineering, Chinese Academy of Sciences. She is currently working at theNational Computer Network Emergency Response Technical Team/Coordination Center of China. Her research interests include knowledgegraph embedding, knowledge acquisition and Web mining.
0000-0002-9660-0291
Bowen Yu
B.W. Yu provided constructive solutions of model components. All the authors have made meaningful and valuable contributions to this manuscript.
Bowen Yu is currently pursuing his PhD degree in the Institute of InformationEngineering, Chinese Academy of Sciences. His research interests includeinformation extraction, metric learning and unsupervised learning.
0000-0002-6804-1859
Lihong Wang
L.H. Wang guided the team of the competition. All the authors have made meaningful and valuable contributions to this manuscript.
wlh@isc.org.cn
Lihong Wang is a Professor at the National Computer Network EmergencyResponse Technical Team/Coordination Center of China. Her current researchinterests include big data mining and analytics, information retrieval andnatural language processing.
0000-0003-0179-2364
Min He
M. He, T.W. Liu and H.B. Xu provided insightful and constructive comments on this manuscript. All the authors have made meaningful and valuable contributions to this manuscript.
Min He is currently working at the National Computer Network EmergencyResponse Technical Team/Coordination Center of China. Her research interestsinclude natural language processing, Web mining and information security.
Tingwen Liu
M. He, T.W. Liu and H.B. Xu provided insightful and constructive comments on this manuscript. All the authors have made meaningful and valuable contributions to this manuscript.
Tingwen Liu is an Associate Researcher of the Institute of InformationEngineering, Chinese Academy of Sciences. His research interests includeinformation extraction and knowledge fusion, text understanding, semanticcomputing, and heterogeneous network representation.
Hongbo Xu
M. He, T.W. Liu and H.B. Xu provided insightful and constructive comments on this manuscript. All the authors have made meaningful and valuable contributions to this manuscript.
Hongbo Xu is a researcher in the Institute of Information Engineering, ChineseAcademy of Sciences. His current research interests include knowledge graphand Web mining.
This work is supported by the National Key Research and Development Program of China (No.2016YFB1000105) and the National Natural Science Foundation of China (No. 61772151). This work’scomputing device is also supported by Beijing Advanced Innovation Center of Big Data and Brain Computing,Beihang University. The author Shu Guo is supported by “Zhizi Program”.
Publication records
Published: Sept. 16, 2021 (Versions1
References
Data Intelligence