Elamur Research⌕
☰
home / papers / 2011 / Chinese Discourse Relation Recognition

Chinese Discourse Relation Recognition

Written for the Research library (claude-opus-5-5) from the paper text, in our own words: the license "CC BY-NC-SA 3.0 (ACL Anthology, before 2016)" does not allow redistributing the text itself. Original — aclanthology.org/I11-1170.pdf

Summary

The paper tackles automatic recognition of discourse relations between consecutive text spans in Chinese. At the time, no Chinese discourse-annotated resource existed, so the authors built one. They sampled 81 articles from Sinica Treebank 3.1 (Sino and Travel sets) and treated each sentence as one argument. Three annotators assigned one of the four top-level PDTB classes (Temporal, Contingency, Comparison, Expansion) to every pair of adjacent sentences, with a fourth annotator breaking ties.

The classifier is a multi-class SVM. Minority classes were oversampled to balance training. Features include argument and clause lengths, final punctuation, a hand-built connective dictionary, shared words and hypernyms, bags of words, POS tags and hypernyms, mined cross-argument word collocations, and number/date indicators. Under 5-fold cross-validation, the full feature set reaches 88.28% accuracy and a macro F-Score of 63.69%. Expansion is recognised best, while Contingency stays weak because the sentence-level scheme misses its typical intra-sentence occurrences.

The work offers an early Chinese baseline and corpus for discourse parsing.

Key points

  • Corpus: 3,081 sentence pairs from 81 Sinica Treebank 3.1 articles (27 Sino, 54 Travel). Labels are heavily skewed: Expansion 84.81%, Comparison 6.72%, Temporal 5.42%, and Contingency 3.05% (only 94 pairs).
  • The connective dictionary contains 319 individual connectives plus 489 paired connectives. The classifier is SVM-multiclass with proportional duplication of the minority classes.
  • Per-class F-Scores with all features: Expansion 93.57%, Temporal 63.46%, Comparison 61.24%, Contingency 36.49%.
  • No single feature works well for all classes. Bag-of-words (71.79% accuracy) and collocated words (70.46%) are the strongest individual features, and combining all features raises accuracy to 88.28%.
  • Treating whole sentences as arguments loses most Contingency cases, which Chinese usually expresses within a single sentence (e.g., 因為…所以…). This is the main limitation the authors identify.

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); 4 of them have a page here, 14 not yet.

Cite

The paper itself, not this page. BibTeX from the library's record of the work.

@inproceedings{huang2011chinese,
  title = {{Chinese Discourse Relation Recognition}},
  author = {Huang, Hen-Hsen and Chen, Hsin-Hsi},
  booktitle = {Proceedings of 5th International Joint Conference on Natural Language Processing},
  year = {2011},
  url = {https://aclanthology.org/I11-1170/}
}

References

7 references, 7 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. L. Carlson, D. Marcu, and M. E. Okurowski. 2001. Building a Discourse-Tagged Corpus in the Framework of Rhetorical Structure Theory. In Proceedings of Second SIGDIAL Workshop on Discourse and Dialogue-Volume 16, pages 1-10. find our card
  2. Hugo Hernault, Danushka Bollegala, and Mitsuru Ishizuka. 2010. A Semi-Supervised Approach to Improve Classification of Infrequent Discourse Relations using Feature Vector Extension. In Proceedings of the 2010 Conference on Empirical Methods in Natural Language Processing, pages 399-409. find our card
  3. 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, pages 343-351. find our card
  4. D. Marcu and A. Echihabi. 2002. An unsupervised Approach to Recognizing Discourse Relations. In Proceedings of the 40th Annual Meeting on Association for Computational Linguistics (ACL 2002), pages 368–375, Morristown, NJ, USA. find our card
  5. Emily Pitler, Annie Louis, and Ani Nenkova. 2009. Automatic Sense Prediction for Implicit Discourse Relations in Text. In Proceedings of the 47th annual Meeting of the ACL and the 4th IJCNLP of the AFNLP, pages 683-691. find our card
  6. Emily Pitler and Ani Nenkova. 2009. Using Syntax to Disambiguate Explicit Discourse Connectives in Text. In Proceedings of the 47th annual Meeting of the ACL and the 4th IJCNLP of the AFNLP. Short Papers. find our card
  7. R. Prasad, N. Dinesh, A. Lee, E. Miltsakaki, L. Robaldo, A. Joshi, and B. Webber. 2008. The Penn Discourse TreeBank 2.0. In Proceedings of LREC’08. find our card