基于指针网络的汉语宏观篇章结构双向解析方法

摘要
参考文献
相关文章 (9)

全文: PDF (6005 KB) HTML (1 KB)
输出: BibTeX | EndNote (RIS)

摘要宏观篇章结构解析旨在通过分析篇章的整体结构,为理解篇章的内容和主旨奠定基础。现有的研究大都采用了单一的自顶向下或自底向上的构建策略逐级地构建结构树,而单向构建策略无法根据不同待解析序列选择合适的解析动作,在解析流程中容易陷入决策局限性并将错误向后传播。该文提出一种集成自顶向下和自底向上两种构建策略的指针网络模型,该模型能同时利用两种构建策略的语义信息,从而选择合适的构建方式。在汉语宏观篇章树库(MCDTB 2.0)上的实验表明,通过集成两种构建方式,该文模型能有效提升篇章单元间的局部语义交互能力并减少构建过程中的错误传播,从而取得性能最优值。

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	何垅旺
	范亚鑫
	褚晓敏
	蒋峰
	李军辉
	李培峰

关键词 ：宏观篇章解析, 结构识别, 指针网络, XLNet

Abstract：Macro discourse structure analysis aims to facilitate the understanding of the content and purpose in a discourse by revealing its structure. This paper proposes a pointer network model integrating top-down and bottom-up construction strategies. It can use the semantic information of the two construction strategies at the same time, so as to select the appropriate construction method. Experiments on Chinese Macro Chinese Discourse Treebank (MCDTB 2.0) show that the model proposed in this paper can effectively reduce the error propagation in the construction process and achieve the best performance.

Key words： macro discourse analysis structure recognition pointer network XLNet

收稿日期: 2021-11-20

中图分类号:

TP391

基金资助:国家自然科学基金(61836007)

作者简介: 何垅旺(1998—),硕士研究生,主要研究领域为自然语言处理。E-mail:20205227115@stu.suda.edu.cn
范亚鑫(1997—),博士研究生,主要研究领域为自然语言处理。E-mail:yxfan@stu.suda.edu.cn
褚晓敏(1981—),通信作者,博士,讲师,主要研究领域为自然语言处理。E-mail:xmchu@suda.edu.cn

链接本文:

http://jcip.cipsc.org.cn/CN/Y2022/V36/I11/68

[1] LIAKATA M, DOBNIK S, Saha S, et al. A discourse-driven content model for summarising scientific articles evaluated in a complex question answering task[C]//Proceedings of the Conference on Empirical Methods in Natural Language Processing, 2013: 747-757.
[2] MEYER T,POPESCU BELIS A. Using sense-labeled discourse connectives for statistical machine translation[C]//Proceedings of the EACL Workshop on Hybrid Approaches to Machine Translation, 2012.
[3] GUZMN F, JOTY S, MRQUEZ L, et al. Using discourse structure improves machine translation evaluation[C]//Proceedings of the 52nd Annual Meeting of the Association for Computational Linguistics, 2014: 687-698.
[4] PELDSZUS A, STEDE M. Joint prediction in mst-style discourse parsing for argumentation mining[C]//Proceedings of the Conference on Empirical Methods in Natural Language Processing, 2015: 938-948.
[5] ATKINSON J, MUNOZ R. Rhetorics-based multi-document summarization[J]. Expert Systems with Applications, 2013, 40(11): 4346-4352.
[6] FERREIRA R, DE SOUZA CABRAL L, FREITAS F, et al. A multi-document summarization system based on statistics and linguistic treatment[J]. Expert Systems with Applications, 2014, 41(13): 5780-5787.
[7] COHAN A,GOHARIAN N. Scientific article summarization using citation-context and article's discourse structure[J]. arXiv preprint arXiv: 1704.06619, 2017.
[8] PRESUTTI V, DRAICCHIO F, GANGEMI A. Knowledge extraction based on discourse representation theory and linguistic frames[C]//Proceedings of the International Conference on Knowledge Engineering and Knowledge Management. Springer, Berlin, Heidelberg, 2012: 114-129.
[9] ZOU B, ZHOU G, ZHU Q. Negation focus identification with contextual discourse information[C]//Proceedings of the 52nd Annual Meeting of the Association for Computational Linguistics, 2014: 522-530.
[10] JIANG F, XU S, CHU X, et al. MCDTB: A macro-level Chinese discourse treebank[C]//Proceedings of the 27th International Conference on Computational Linguistics, 2018: 3493-3504.
[11] FAN Y, JIANG F, CHU X, et al. Combining global and local information to recognize Chinese macro discourse structure [C]//Proceedings of the 19th Chinese National Conference on Computational Linguistics, 2020: 183-194.
[12] LIU L, LIN X,JOTY S, et al. Hierarchical pointer net parsing[J]. arXiv preprint arXiv: 1908.11571, 2019.
[13] LIN X, JOTY S, JWALAPURAM P, et al. A unified linear-time framework for sentence-level discourse parsing[OL]https://www.aclweb.org/anthology/p19-1400.
[14] KOTO F, LAU J H, BALDWIN T. Top-down discourse parsing via sequence labelling[J]. arXiv preprint arXiv: 2102.02080, 2021.
[15] ZHOU Y, CHU X, LI P, et al. Constructing Chinese macro discourse tree via multiple views and word pair similarity[C]//Proceedings of the CCF International Conference on Natural Language Processing and Chinese Computing. Springer, Cham, 2019: 773-786.
[16] JIANG F, CHU X, LI P, et al. Chinese paragraph-level discourse parsing with global backward and local reverse reading[C]//Proceedings of the 28th International Conference on Computational Linguistics, 2020: 5749-5759.
[17] JIANG F, FAN Y, CHU X, et al. Hierarchical macro discourse parsing based on topic segmentation[C]//Proceedings of the AAAI Conference on Artificial Intelligence, 2021, 35(14): 13152-13160.
[18] FENG V W, HIRST G. A linear-time bottom-up discourse parser with constraints and post-editing[C]//Proceedings of the 52nd Annual Meeting of the Association for Computational Linguistics, 2014: 511-521.
[19] LI Q, LI T, CHANG B. Discourse parsing with attention-based hierarchical neural networks[C]//Proceedings of the Conference on Empirical Methods in Natural Language Processing, 2016: 362-371.
[20] KOBAYASHI N,HIRAO T, NAKAMURA K, et al. Split or merge: Which is better for unsupervised RST parsing?[C]//Proceedings of the Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing, 2019: 5797-5802.
[21] CARLSON L,MARCU D, OKUROWSKI M E. Building a discourse-tagged corpus in the framework of rhetorical structure theory[M]. Current and New Directions in Discourse and Dialogue. Springer, Dordrecht, 2003: 85-112.
[22] MABONA A, RIMELL L, CLARK S, et al. Neural generative rhetorical structure parsing[J]. arXiv preprint arXiv: 1909.11049, 2019.
[23] FRIED D, STERN M, KLEIN D. Improving neural parsing by disentangling model combination and reranking effects[J].arXiv preprint arXiv: 1707.03058, 2017.
[24] ZHANG L, TAN X, KONG F, et al. A recursive information flow gated model for RST-style text-level discourse parsing[C]//Proceedings of the CCF International Conference on Natural Language Processing and Chinese Computing. Springer, Cham, 2019: 231-241.
[25] KOBAYASHI N,HIRAO T, KAMIGAITO H, et al. Top-down RST parsing utilizing granularity levels in documents[C]//Proceedings of the AAAI Conference on Artificial Intelligence, 2020, 34(05): 8099-8106.
[26] NGUYEN T T, NGUYEN X P, JOTY S, et al. RST parsing from scratch[J]. arXiv preprint arXiv: 2105.10861, 2021.
[27] VINYALS O, FORTUNATO M, JAITLY N. Pointer networks[J]. arXiv preprint arXiv: 1506.03134, 2015.
[28] CUI Y, CHE W, LIU T, et al. Revisiting pre-trained models for Chinese natural language processing[J]. arXiv preprint arXiv: 2004.13922, 2020.
[29] CHO K, VANMERRINBOER B, GULCEHRE C, et al. Learning phrase representations using RNN encoder-decoder for statistical machine translation[J]. arXiv preprint arXiv: 1406.1078, 2014.
[30] HOCHREITER S, SCHMIDHUBER J. Long short-term memory[J]. Neural Computation, 1997, 9(8): 1735-1780.