Elamur ResearchβŒ•
☰
home / papers / 2018 / Using active learning to expand training data for implicit discourse relation recognition

Using active learning to expand training data for implicit discourse relation recognition

Written for the Research library (claude-opus-5-5) from the paper text. Original β€” aclanthology.org/D18-1079.pdf

Summary

Implicit discourse relations, where no connective such as "because" signals how two arguments relate, are hard to classify. Neural models need more labeled data than PDTB v2.0 provides. Following Rutherford and Xue (2015), the authors create extra training pairs by taking explicitly connected arguments, labeling each pair with the relation its connective most often signals, and deleting the connective. Adding all of these artificial pairs actually hurt performance. The authors therefore added an active learning loop: a Siamese CNN (a simplified version of Qin et al. 2016) is trained on real implicit pairs and scores the explicit candidates by prediction entropy. Only the most uncertain candidates are labeled by the Exp2Imp module and added to the training set. On four-way classification of top-level PDTB relations, this targeted expansion raised both accuracy and macro F1. The weak CNN then rivaled more complex systems that rely on larger external resources. This suggests that careful selection of synthetic data matters more than its volume.

Key points

  • Data split on PDTB v2.0: sections 2–20 for training (the labeled pool), 21–22 for testing, and 0–1 for development. A set of 450 explicit instances was used to map connectives to relation types.
  • Informativeness is the entropy of the CNN's class probabilities. Candidates scoring above a threshold ΞΈ = 0.95 are selected in independent batches.
  • Using roughly 15% of the unlabeled pool, AL raised accuracy from 56.78 to 60.63 (+3.85) and macro F1 from 40.85 to 44.48 (+3.63).
  • The "Blender" approach, which added every converted pair, reduced accuracy by 2.08 and macro F1 by 1.73 points.
  • Compared with IO, MNN, MTN, DSWE and MANN, the method achieved the best accuracy with comparable F1. Code and selected data are available at github.com/AndreaXu0401/ALIDRC.

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

Fields: Linguistics & languages

Topics: Discourse analysis

Cited by

11 papers of our library cite it (of the 140673 whose bibliographies we read); 7 of them have a page here, 4 not yet.

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/d18-1079

@inproceedings{xu2018using,
  title = {{Using active learning to expand training data for implicit discourse relation recognition}},
  author = {Xu, Yang and Hong, Yu and Ruan, Huibin and Yao, Jianmin and Zhang, Min and Zhou, Guodong},
  booktitle = {Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing},
  year = {2018},
  doi = {10.18653/v1/d18-1079},
  url = {https://doi.org/10.18653/v1/d18-1079}
}

References

29 references, 20 of them with a card of ours. β†— leads to the work, Β«findΒ» searches for one we have no id for yet.

  1. Annie Louis, Aravind Joshi, Rashmi Prasad, and Ani Nenkova. 2010. Using entity features to classify implicit discourse relations. In Proceedings of the 11th Annual Meeting of the Special Interest Group on Discourse and Dialogue, pages 59–62. Association for Computational Linguistics. find our card
  2. Hongxiao Bai and Hai Zhao. 2018. Deep enhanced representation for implicit discourse relation recognition. In Proceedings of the 27th International Conference on Computational Linguistics, pages 571–583. Association for Computational Linguistics. find our card
  3. Tomas Mikolov, Kai Chen, Greg Corrado, and Jeffrey Dean. 2013. Efficient estimation of word representations in vector space. arXiv preprint arXiv:1301.3781. find
  4. Chloé Braud and Pascal Denis. 2016. Learning connective-based word representations for implicit discourse relation identification. In EMNLP, pages 203–213. find our card
  5. Joonsuk Park and Claire Cardie. 2012. Improving implicit discourse relation recognition through feature set optimization. In Proceedings of the 13th Annual Meeting of the Special Interest Group on Discourse and Dialogue, pages 108–112. Association for Computational Linguistics. find our card
  6. Lynn Carlson, Daniel Marcu, and Mary Ellen Okurowski. 2003. Building a discourse-tagged corpus in the framework of rhetorical structure theory. In Current and new directions in discourse and dialogue, pages 85–112. Springer. find our card
  7. 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
  8. Jifan Chen, Qi Zhang, Pengfei Liu, Xipeng Qiu, and Xuanjing Huang. 2016. Implicit discourse relation detection via a deep architecture with gated relevance network. In ACL (1). find our card
  9. Rashmi Prasad, Nikhil Dinesh, Alan Lee, Eleni Milt-sakaki, Livio Robaldo, Aravind K Joshi, and Bonnie L Webber. 2008. The penn discourse treebank 2.0. In LREC. Citeseer. find our card
  10. Fengyu Guo, Ruifang He, Di Jin, Jianwu Dang, Long-biao Wang, and Xiangang Li. 2018. Implicit discourse relation recognition using neural tensor network with interactive attention and sparse learning. In Proceedings of the 27th International Conference on Computational Linguistics, pages 547–558. Ass… find our card
  11. Lianhui Qin, Zhisong Zhang, and Hai Zhao. 2016. A stacking gated neural architecture for implicit discourse relation classification. In Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing, pages 2263–2270. find our card
  12. Husheng Guo and Wenjian Wang. 2015. An active learning-based svm multi-class classification model. Pattern recognition, 48(5):1577–1597. find
  13. Lianhui Qin, Zhisong Zhang, Hai Zhao, Zhiting Hu, and Eric P Xing. 2017. Adversarial connective-exploiting networks for implicit discourse relation classification. arXiv preprint arXiv:1704.00217. find our card
  14. Man Lan, Jianxiang Wang, Yuanbin Wu, Zheng-Yu Niu, and Haifeng Wang. 2017. Multi-task attention-based neural networks for implicit discourse relationship representation and identification. In Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing, pages 1299– 1308. find our card
  15. Maria E Ramirez-Loaiza, Manali Sharma, Geet Kumar, and Mustafa Bilgic. 2017. Active learning: an empirical study of common baselines. Data Mining and Knowledge Discovery, 31(2):287–313. find
  16. Xin Li and Yuhong Guo. 2013. Active learning with multi-label svm classification. In IJCAI, pages 1479–1485. find
  17. Attapol Rutherford and Nianwen Xue. 2014. Discovering implicit discourse relations through brown cluster pair representation and coreference patterns. In EACL, volume 645, page 2014. find our card
  18. 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
  19. Attapol Rutherford and Nianwen Xue. 2015. Improving the inference of implicit discourse relations via classifying explicit discourse connectives. In HLT-NAACL, pages 799–808. find our card
  20. Evan Sandhaus. 2008. The new york times annotated corpus. Linguistic Data Consortium, Philadelphia, 6(12):e26752. find
  21. Yang Liu and Sujian Li. 2016. Recognizing implicit discourse relations via repeated reading: Neural networks with multi-level attention. arXiv preprint arXiv:1609.06380. find our card
  22. Burr Settles. 2010. Active learning literature survey. University of Wisconsin, Madison, 52(55-66):11. find our card
  23. Yang Liu, Sujian Li, Xiaodong Zhang, and Zhifang Sui. 2016. Implicit discourse relation classification via multi-task neural networks. In AAAI, pages 2750– 2756. find
  24. Changxing Wu, Yidong Chen, Yanzhou Huang, et al. 2016. Bilingually-constrained synthetic data for implicit discourse relation recognition. In Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing, pages 2306–2312. find our card
  25. Changxing Wu, Xiaodong Shi, Yidong Chen, Jinsong Su, and Boli Wang. 2017. Improving implicit discourse relation recognition with discourse-specific word embeddings. In Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers), pages 269– 274, Va… find our card
  26. Yi Yang, Zhigang Ma, Feiping Nie, Xiaojun Chang, and Alexander G Hauptmann. 2015. Multi-class active learning by uncertainty sampling with diversity maximization. International Journal of Computer Vision, 113(2):113–127. find
  27. Biao Zhang, Jinsong Su, Deyi Xiong, Yaojie Lu, Hong Duan, and Junfeng Yao. 2015. Shallow convolutional neural network for implicit discourse relation recognition. In EMNLP, pages 2230–2235. find our card
  28. Ye Zhang, Matthew Lease, and Byron C Wallace. 2017. Active discriminative text representation learning. In AAAI, pages 3386–3392. find
  29. Jingbo Zhu, Huizhen Wang, Tianshun Yao, and Benjamin K Tsou. 2008. Active learning with sampling by uncertainty and density for word sense disambiguation and text classification. In Proceedings of the 22nd International Conference on Computational Linguistics-Volume 1, pages 1137–1144. Association f… find

↑ back to top