Abstract
Document-level event extraction aims to extract event-related information from an unstructured document composed of multiple sentences. Existing approaches are not effective due to the challenge of event arguments that are scattered across multi-sentences and they pay more attention to the coreference relationship between entity mentions. However, it is an extremely common phenomenon that there are a large number of crossing sentences pronouns that referring to entity mentions. These pronouns also contain rich semantic information related to events in the document. Therefore, there is still a challenge that how to effectively construct the mention–pronoun coreference relationship and better learn the rich semantic entities representations for DEE. Aiming at the above problems, we propose a novel document-level multi-task learning approach based on coreference-aware dynamic heterogeneous graph network for event extraction, named DMCGEE. Specifically, first, an information enhancement extractor module is constructed to effectively capture multi-types of semantic association information for mentions representations. Second, a mention–pronoun coreference resolution method is proposed to capture mention–pronoun coreference resolution pairs, and a coreference-aware dynamic heterogeneous graph network is constructed to help sentences and mentions representations to focus on the effective global related information, thereby improving the performance of DMCGEE. Experiments show that DMCGEE outperforms the state-of-the-art.
Similar content being viewed by others
Data Availability
The open datasets analyzed during this study have been properly cited in this published article (see reference section). If found difficulty in finding the data links, same can be available from the corresponding author on reasonable request. Our own dataset has been uploaded to Github, it can be accessed through https://github.com/just123cz/Paper_data.git.
References
J H, Li S, Ji H (2021) Document-level event argument extraction by conditional generation, arXiv preprint arXiv:2104.05919
Y C, Liu J, (2020) Event extraction as machine reading comprehension, EMNLP 1641–1651
W C, Zheng S, (2019) Doc2edag: An end-to-end document-level framework for chinese financial event extraction, EMNLP, pp 337–346
Fei H, Ren Y (2021) Enriching contextualized language model from knowledge graph for biomedical information extraction. Brief Bioinform 23:bba110
Sarrouti M, En-Nahnahi N (2021) Mttlade: a multi-task transfer learning-based method for adverse drug events extraction. Inform Process Manag 58:102473
O S, Tunaoglu D, (2009) Event extraction from turkish football web-casting texts using hand-crafted templates. In: IEEE international conference on semantic computing, pp 466–472
J D, Cybulka J, (2015) Events extractor for polish based on semantics-driven extraction templates. In: language and technology conference, pp 231–245
S L, Sha L, (2015) Joint learning templates and slots for event schema induction, NAACL pp 428–434
E M, Olubanjo T, (2016) Detecting food intake acoustic events in noisy recordings using template matching. In: IEEE-EMBS international conference on biomedical and health informatics. pp 388–391
L T, Runxin Xu, (2021) Document-level event extraction via heterogeneous graph-based interaction model with a tracker, ACL/IJCNP, pp 3533–3546
HangYang S D. (2021) Document-level event extraction via parallel prediction networks, ACL/IJCNP pp 6298–6308
D X. NingLuo (2021) A framework for document-level cybersecurity event extraction from open source data, CSCWD, pp 422–427
Xinya CC (2021) Grit: Generative role-filler transformers for document-level event entity extraction, EACL 634–644
Yaojie Liu LH (2021) Text2event: Controllable sequence-to-structure generation for end-to-end event extraction, ACL/IJCNP pp 2795–2806
J H. Ying Lin (2020) A joint neural model for information extraction with global features, ACL pp 5929–5939
David Wadden UW (2019) Entity, relation, and event extraction with contextualized span representations, EMNLP, pp 5783–5788
e. a. Wasi Uddin Ahmad, Nanyun Peng (2021) Gate: graph attention transformer encoder for cross-lingual relation and event extraction, AAAI, pp 12462–12470
O M, CaseLLi T (2021) Chinese ner using lattice lstm, Proceedings of the 4th workshop on challenges and applications of automated extraction of socio-political events from text, pp 12–19
Yang HYC (2018) Dcfee: A document-level chinese financial event extraction system based on automatically labeled training data, Proceedings of ACL, pp 50–55
Liu XH (2019) Open domain event extraction using neural latent variable models, Proceedings of ACL 2019 2860–2871
Du XCC (2019) Document-level event role filler extraction using multi granularity contextualized encoding. Proceedings of ACL 2020, pp 2860–2871
Huang RER (2012) Modeling textual cohesion for event extraction, AAAI 2012, pp 2860–2871
Hang Yang YC (2018) Dcfee: A document-level chinese financial event extraction system based on automatically labeled training data, ACL 2018, pp 50–55
Yusheng Huang WJ (2021) Exploring sentence community for document-level event extraction, EMNLP 2021, pp 340–351
Krupka GR,(1995) Description of the sra system as used for muc-6, Proceedings of the 6th Message understanding Conference, pp 221–235
Shaalan K (2009) Nera: named entity recognition for Arabic. JAm Soc 60:1652–1663
Bikel DM (1999) An algorithm that learns what’s in a name, Machine learning, pp 211–231
ISOZAKIH H (2002) Efficient support vector classifiers for named entity recognition, ACL 1–7
Collobert R, Weston J (2011) Natural language processing (almost) from scratch. J Mach Learn Res 12:2493–2537
Yao L, Liu H, Liu Y, Li X (2015) Biomedical named entity recognition based on deep neutral network. Int J Hybrid Inform Technol 8:279–288
Zhang Y, Yang J (2018) Chinese ner using lattice lstm. In: proceedings of the 56th international conference on computational linguistics, pp 1554–1564
Sun K, Zhang R, Mensah S, Mao Y, Liu X (2021) Progressive multi-task learning with controlled information flow for joint entity and relation extraction. AAAI 35:13851–13859
Ya Xiao CT (2020) Joint entity and relation extraction with a hybrid transformer and reinforcement learning based model. AAAI 2020:9314–9321
Yan S, Lin KJ, Zheng X, Wang H (2021) LkeRec: toward lightweight end-to-end joint representation learning for building accurate and effective recommendation. ACM Trans Inf Syst 40(3):1–28
Wang Y, Liu H (2020) Hybrid neural recommendation with joint deep representation learning of ratings and reviews. Neurocomputing 374:77–85
Li S, Ma Q (2020) Joint-label learning by dual augmentation for time series classification, IEEE Transactions on Pattern Analysis and Machine Intelligence, pp 77–85
Z Z, Ma Q (2021) Joint-label learning by dual augmentation for time series classification, AAAI 2021, pp 8847–8855
Zhang H, Bai J (2021) Joint coreference resolution and character linking for multiparty conversation, EACL, pp 539–548
Huang YJ, Kurohashi S (2021) Extractive summarization considering discourse and coreference relations based on heterogeneous graph, EACL, pp 3046–3052
L P, Zeyu Dai, Hongliang Fei (2019) Coreference aware representation learning for neural named entity recognition. In: proceedings of the twenty-eighth international joint conference on artificial intelligence main track, pp 4946–4953
T D, Zaporojets K, Deleu J (2021) Towards consistent document-level entity linking: Joint models for entity linking and coreference resolution
Q D, Xue Z, Li R (2022) Corefdre: Document-level relation extraction with coreference resolution
H R, Prafulla Kumar Choubey (2017) A sequential model for classifying temporal relations between intra-sentence events, EMNLP, pp 1796–1802
William AS, Mann C (1988) Rhetorical structure theory: toward a functional theory of text organization. Text-Interdiscip J Study Dis 27:243–281
Jason Weston SC, Memory networks, ICIR (2015)
Kendall A, Y Gal (2018) Multi-task learning using uncertainty to weigh losses for scene geometry and semantics. In: proceedings of the IEEE conference on computer vision and pattern recognition, pp 7482–7491
Sha L, Heng J (2021) Document-level event argument extraction by conditional generation, NAACL, pp 894-908
Ying L (2020) A joint neural model for information extraction with global features, ACL, pp 7999-8009
Xinya D (2020) Event extraction by answering (almost) natural questions, EMNLP, pp 671–683
Acknowledgements
This research is funded by the Applied Basic Research Program of Liaoning Province (No. 2022JH2/101300250). Digital Liaoning Smart Building Strong Province (Direction of Digital Economy) (No. 13031307053000568). National Natural Science Foundation of China (No. 62072220, 61502215). Central Government Guides Local Science and Technology Development Foundation Project of Liaoning Province (No. 2022JH6/100100032). Natural Science Foundation of Liaoning Province (2022-KF-13-06). National Natural Science Foundation of China (61472169). The youth talent support program of ‘Xing Liao Talent Program’ (No. XLYC2203003).
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no conflict of interests in this work.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Chen, Z., Ji, W., Ding, L. et al. Document-level multi-task learning approach based on coreference-aware dynamic heterogeneous graph network for event extraction. Neural Comput & Applic 36, 303–321 (2024). https://doi.org/10.1007/s00521-023-08977-0
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00521-023-08977-0