Elamur Research⌕
☰
home / papers / 2016 / Two End-to-end Shallow Discourse Parsers for English and Chinese in CoNLL-2016 Shared Task

Two End-to-end Shallow Discourse Parsers for English and Chinese in CoNLL-2016 Shared Task

Summary

The paper presents two end-to-end shallow discourse parsers, one for English and one for Chinese, submitted to the CoNLL-2016 Shared Task. The parsers identify discourse connectives, extract their two arguments and assign relation senses. The English system extends the authors' CoNLL-2015 winning pipeline (Wang and Lan, 2015b) in three ways. It handles Arg1 and Arg2 inside a single sentence with two distinct constituent-level binary classifiers. It replaces handcrafted features for Non-Explicit sense labelling with a CNN over word2vec embeddings. It re-tunes each component's features through hill-climbing. The Chinese system is a new pipeline that mirrors how the Chinese Discourse Treebank was annotated. It relies on punctuation-based chunking and accounts for paired connectives and sense-dependent argument order. Other components use logistic regression in LIBLINEAR. On the blind test data, the English parser outperformed the previous year's best system, and both parsers placed second. The work shows that Chinese needs a parsing design that reflects its own discourse properties rather than a direct port of English methods.

Key points

  • English overall parser F1 on the blind set (English Wikinews) was 25.99, compared with 24.00 for the CoNLL-2015 best system. On the PDTB dev and test sets it scored 40.43 and 30.70.
  • CNN setup: 300-dimensional Google News word2vec vectors, filter windows (4, 6, 13), 100 feature maps, Tanh, max pooling, dropout 0.5, AdaGrad (learning rate 0.01, minibatch 50).
  • With gold arguments, the CNN raised English Non-Explicit sense accuracy on test from 34.45 to 40.91. On the blind set it fell behind the old system (34.20 vs 36.29), which the authors attribute to a different sense distribution.
  • About 62.46% of Chinese training connectives are paired. Chinese Explicit connective detection fell from 86.27 (dev) to 63.07 (blind), and the overall Chinese blind score was 26.60.
  • Chinese Non-Explicit sense accuracy is high (60.52 blind) partly because labelling every relation as "Conjunction" already reaches 64.61% on the training set.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

Fields: Linguistics & languages

Topics: Discourse analysis

Cited by

18 papers of our library cite it (of the 140673 whose bibliographies we read); 11 of them have a page here, 7 not yet.

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/k16-2004

@inproceedings{wang2016two,
  title = {{Two End-to-end Shallow Discourse Parsers for English and Chinese in CoNLL-2016 Shared Task}},
  author = {Wang, Jianxiang and Lan, Man},
  booktitle = {Proceedings of the CoNLL-16 shared task},
  year = {2016},
  doi = {10.18653/v1/k16-2004},
  url = {https://doi.org/10.18653/v1/k16-2004}
}

References

21 references, 12 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Chloé Braud and Pascal Denis. 2015. Comparing word representations for implicit discourse relation classification. In Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing, pages 2201–2211, Lisbon, Portugal, September. Association for Computational Linguistics. find our card
  2. John Duchi, Elad Hazan, and Yoram Singer. 2011. Adaptive subgradient methods for online learning and stochastic optimization. The Journal of Machine Learning Research, 12:2121–2159. find
  3. Robert Fisher and Reid Simmons. 2015. Spectral semi-supervised discourse relation classification. In Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing (Volume 2: Short Papers), pages 89–93… find our card
  4. Geoffrey E Hinton, Nitish Srivastava, Alex Krizhevsky, Ilya Sutskever, and Ruslan R Salakhutdinov. 2012. Improving neural networks by preventing co-adaptation of feature detectors. arXiv preprint arXiv:1207.0580. find
  5. Yangfeng Ji and Jacob Eisenstein. 2015. One vector is not enough: Entity-augmented distributed semantics for discourse relations. Transactions of the Association for Computational Linguistics, 3:329–344. find our card
  6. Fang Kong, Hwee Tou Ng, and Guodong Zhou. 2014. A constituent-based approach to argument labeling with joint inference in discourse parsing. In Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 68–77, Doha, Qatar, October. Association for Computati… find our card
  7. Man Lan, Yu Xu, Zheng-Yu Niu, et al. 2013. Leveraging synthetic discourse data via multi-task learning for implicit discourse relation recognition. In ACL (1), pages 476–485. Citeseer. find our card
  8. Ziheng Lin, Min-Yen Kan, and Hwee Tou Ng. 2009. Recognizing implicit discourse relations in the Penn Discourse Treebank. In Proceedings of the 2009 Conference on Empirical Methods in Natural Language Processing: Volume 1-Volume 1, pages 343–351. Association for Computational Linguistics. find our card
  9. Ziheng Lin, Hwee Tou Ng, and Min-Yen Kan. 2014. A PDTB-styled end-to-end discourse parser. Natural Language Engineering, pages 1– 34. find
  10. Tomas Mikolov, Kai Chen, Greg Corrado, and Jeffrey Dean. 2013. Efficient estimation of word representations in vector space. arXiv preprint arXiv:1301.3781. find
  11. Emily Pitler and Ani Nenkova. 2009. Using syntax to disambiguate explicit discourse connectives in text. In Proceedings of the ACL-IJCNLP 2009 Conference Short Papers, pages 13–16. Association for Computational Linguistics. find our card
  12. Shared Task on Shallow Discourse Parsing. In Proceedings of the Nineteenth Conference on Computational Natural Language Learning: Shared Task, Beijing, China. find
  13. Nianwen Xue, Hwee Tou Ng, Sameer Pradhan, Bonnie Webber, Attapol Rutherford, Chuan Wang, and Hongmin Wang. 2016. The conll-2016 shared task on multilingual shallow discourse parsing. In Proceedings of the Twentieth Conference on Computational Natural Language Learning - Shared Task, Berlin, Germany,… find our card
  14. Biao Zhang, Jinsong Su, Deyi Xiong, Yaojie Lu, Hong Duan, and Junfeng Yao. 2015. Shallow convolutional neural network for implicit discourse relation recognition. In Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing, pages 2230–2235, Lisbon, Portugal, September.… find our card
  15. Yuping Zhou and Nianwen Xue. 2012. Pdtb-style discourse annotation of chinese text. In Proceedings of the 50th Annual Meeting of the Association for Computational Linguistics: Long Papers-Volume 1, pages 69–77. Association for Computational Linguistics. find
  16. Yuping Zhou and Nianwen Xue. 2015. The chinese discourse treebank: A chinese corpus annotated with discourse relations. Language Resources and Evaluation, 49(2):397–431. find
  17. Emily Pitler, Annie Louis, and Ani Nenkova. 2009. Automatic sense prediction for implicit discourse relations in text. In Proceedings of the Joint Conference of the 47th Annual Meeting of the ACL and the 4th International Joint Conference on Natural Language Processing of the AFNLP: Volume 2-Volume… find our card
  18. Attapol T Rutherford and Nianwen Xue. 2014. Discovering implicit discourse relations through brown cluster pair representation and coreference patterns. find our card
  19. Jianxiang Wang and Man Lan. 2015a. Building a high performance end-to-end explicit discourse parser for practical application. In Knowledge Science, Engineering and Management, pages 324–335. Springer. find
  20. Jianxiang Wang and Man Lan. 2015b. A refined end-to-end discourse parser. In Proceedings of the Nineteenth Conference on Computational Natural Language Learning - Shared Task, pages 17–24, Beijing, China, July. Association for Computational Linguistics. find our card
  21. Nianwen Xue, Hwee Tou Ng, Sameer Pradhan, Rashmi Prasad, Christopher Bryant, and Attapol Rutherford. 2015. The CoNLL-2015 find

↑ back to top

This page in Markdown →