Argumentation Mining Based on Multi-task Joint Learning
LIAO Xiangwen1,2,3, NI Jichang1,2,3, WEI Jingjing4, WU Yunbing1,2,3, CHEN Guolong1,2,3
1.College of Mathematics and Computer Science, Fuzhou University, Fuzhou 350116; 2.Fujian Provincial Key Laboratory of Networking Computing and Intelligent Information Processing, Fuzhou University,Fuzhou 350116; 3.Digital Fujian Institute of Financial Big Data, Fuzhou 350116; 4.College of Electronics and Information Science, Fujian Jiangxia University, Fuzhou 350108
Abstract：Most of the existing research on argumentation mining is focused on modeling single dataset, and the possible changes in feature of different datasets are neglected. And thus the generalization performance of the model is decreased. Therefore, an argumentation mining method based on multi-task learning is proposed to combine the argumentation mining tasks of multiple datasets for joint learning. Firstly, the input layers of multiple tasks are fused, and the sharing parameters of word level and character level are obtained via deep convolutional neural network and highway network. The joint task-related feature input into stacking long-short term memory is utilized to train the correlation information between multiple argumentation mining tasks in parallel. Finally, the results of sequence labeling are obtained by the conditional random field. The experimental results on six datasets of various fields verify the effectiveness of the proposed method with increased Macro-F1.
[1] STAB C, GUREVYCH I. Identifying Argumentative Discourse Structures in Persuasive Essays // Proc of the Conference on Empirical Methods in Natural Language Processing. Stroudsburg, USA: ACL, 2014: 46-56. [2] HABERNAL I, GUREVYCH I. Argumentation Mining in User-Ge-nerated Web Discourse. Computational Linguistics, 2017, 43(1): 125-179. [3] MOENS M F, BOIY E, PALAU R M, et al. Automatic Detection of Arguments in Legal Texts // Proc of the International Conference on Artificial Intelligence and Law. New York, USA: ACM, 2007: 225-230. [4] EGER S, DAXENBERGER J, GUREVYCH I. Neural End-to-End Learning for Computational Argumentation Mining // Proc of the 55th Annual Meeting of the Association for Computational Linguistics. Stroudsburg, USA: ACL, 2017: 11-22. [5] FLOROU E, KONSTANTOPOULOS S, KOUKOURIKOS A, et al. Argument Extraction for Supporting Public Policy Formulation // Proc of the 7th Workshop on Language Technology for Cultural He-ritage, Social Sciences, and Humanities. Stroudsburg, USA: ACL, 2013: 49-54. [6] STAB C, GUREVYCH I. Parsing Argumentation Structures in Per-suasive Essays. Computational Linguistics, 2017, 43(3): 619-659. [7] LI M L, GAO Y, WEN H, et al. Joint RNN Model for Argument Component Boundary Detection // Proc of the IEEE International Conference on Systems, Man, and Cybernetics. Washington, USA: IEEE, 2017: 57-62. [8] LAHA A, RAYKAR V. An Empirical Evaluation of Various Deep Learning Architectures for Bi-sequence Classification Tasks // Proc of the 26th International Conference on Computational Linguistics. Berlin, Germany: Springer, 2016: 2762-2773. [9] PERSING I, NG V. End-to-End Argumentation Mining in Student Essays // Proc of the North American Chapter of the Association for Computational Linguistics(Human Language Technologies). Strouds-burg, USA: ACL, 2016: 1384-1394. [10] POTASH P, ALEXEY R, ANNA R. Here′s My Point: Argumen-tation Mining with Pointer Networks[C/OL]. [2019-05-26]. https://arxiv.org/pdf/1612.08994v1.pdf. [11] SCHULZ C, EGER S, DAXENBERGER J, et al. Multi-Task Learning for Argumentation Mining in Low-Resource Settings // Proc of the North American Chapter of the Association for Computational Linguistics(Human Language Technologies). Stroudsburg, USA: ACL, 2018: 35-41. [12] 廖祥文,陈泽泽,桂 林,等.基于多任务迭代学习的论辩挖掘方法.计算机学报, 2019, 42(7): 1524-1538. (LIAO X W, CHENG Z Z, GUI L, et al. An Argumentation Mining Method Based on Multi-task Iterative Learning. Chinese Journal of Computers, 2019, 42(7): 1524-1538.) [13] KIM Y. Convolutional Neural Networks for Sentence Classification // Proc of the Conference on Empirical Methods in Natural Language Processing. Stroudsburg, USA: ACL, 2014: 1746-1751. [14] KIM Y, JERNITE Y, SPNTAG D, et al. Character-Aware Neural Language Models // Proc of the 30th AAAI Conference on Artificial Intelligence. Palo Alto, USA: AAAI Press, 2016: 2741-2749. [15] YANG J, LIANG S L, ZHANG Y. Design Challenges and Mis-conceptions in Neural Sequence Labeling // Proc of the 27th International Conference on Computational Linguistics. Stroudsburg, USA: ACL, 2018: 3879-3889. [16] GAO Y, WANG H, ZHANG C, et al. Reinforcement Learning Based Argument Component Detection[C/OL]. [2019-05-26]. https://arxiv.org/pdf/1702.06239.pdf. [17] ROONEY N, WANG H, FIONA B. Applying Kernel Methods to Argumentation Mining // Proc of the 25th International Florida Artificial Intelligence Research Society Conference. Berlin, Germany: Springer, 2012: 272-275. [18] PELDSZUS A, STEDE M. An Annotated Corpus of Argumentative Microtexts // Proc of the 1st Conference on Argumentation. Berlin, Germany: Springer, 2015: 801-815. [19] PALAU R M, MOENS M F. Argumentation Mining: The Detec-tion, Classification and Structure of Arguments in Text // Proc of the 12th International Conference on Artificial Intelligence and Law. New York, USA: ACM, 2009: 98-107. [20] LEVY R, YONATAN B, DANIEL H, et al. Context Dependent Claim Detection // Proc of the 25th International Conference on Computational Linguistics(Technical Papers). Stroudsburg, USA: ACL, 2014: 1489-1500. [21] DAXENBERGER J, EGER S, HABERNAL I, et al. What is the Essence of a Claim? Cross-Domain Claim Identification // Proc of the Conference on Empirical Method in Natural Language Proce-ssing. Stroudsburg, USA: ACL, 2017: 2045-2056. [22] BINGEL J, SGAARD A. Identifying Beneficial Task Relations for Multi-task Learning in Deep Neural Networks // Proc of the 15th Conference of the European Chapter of the Association for Computational Linguistics. Stroudsburg, USA: ACL, 2017, II: 164-169. [23] BENTON A, MITCHELL M, HOVY D. Multi-task Learning for Mental Health Using Social Media Text // Proc of the 15th Confe-rence of the European Chapter of the Association for Computational Linguistics. Stroudsburg, USA: ACL, 2017, I: 152-162. [24] BRAUD C, PLANK B, SGAARD A. Multi-view and Multi-task Training of RST Discourse Parsers // Proc of the 26th International Conference on Computational Linguistics(Technical Papers). Stroudsburg, USA: ACL, 2016: 1903-1913. [25] REI M, CRICHTON G, PYYSALO S. Attending to Characters in Neural Sequence Labeling Models // Proc of the 26th International Conference on Computational Linguistics(Technical Papers). Stroudsburg, USA: ACL, 2016: 309-318.
2019, Vol. 32 
