Elamur Research⌕
☰
home / papers / 2006 / Classification of Discourse Coherence Relations: An Exploratory Study using Multiple Knowledge Sources

Classification of Discourse Coherence Relations: An Exploratory Study using Multiple Knowledge Sources

Summary

This study tackles two tasks on the Discourse GraphBank (Wolf and Gibson, 2005), a corpus of 135 news articles. Its discourse relations form graphs rather than trees, because segments can have several parents and links can cross. The tasks are (1) labelling the type of coherence relation between two segments and (2) deciding whether two segments are related at all. Gold discourse segments are assumed. Instead of a compositional discourse parser, the authors use a flat Maximum Entropy classifier. It combines many knowledge sources:

  • RASP grammatical relations
  • Carafe CRF tagging and chunking
  • EvITA event features, SlinkET modal subordination and temporal links (from the TARSQI framework)
  • Word Sketch Engine similarity computed over the British National Corpus
  • the Brandeis Semantic Ontology (BSO)

The classifier reaches about 81% accuracy on relation typing. On identifying relations within a single sentence it reaches about 70%. The authors also argue that GraphBank's cause-effect label wrongly merges true causation with purpose and rationale clauses. The work shows that a simple feature-based model can absorb diverse linguistic resources, and its feature analysis can guide more structured approaches such as SDRT or D-LTAG.

Key points

  • Data: Discourse GraphBank, 135 documents, with 11 relation labels. The most frequent are elab, attr, par, same and ce.
  • Relation typing: 81.06% fine-grained and 87.51% coarse-grained accuracy using all feature classes.
  • Per-relation results: attr and elab score F above 91. Rare relations perform poorly; for example, temp scores about 29 F.
  • Ablation: Removing single feature classes shows that some contribute far more than others. Dropping the "P" class lowers accuracy to 71.52%.
  • Relation identification (same-sentence pairs): 70.04% accuracy, and 64.53% when identification and typing are done jointly. The baseline is 58%.
  • Dependency features: Adding features from neighbouring gold relations raises typing accuracy to 82.3%.

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

25 papers of our library cite it (of the 140673 whose bibliographies we read); 13 of them have a page here, 12 not yet.

Cite

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

@inproceedings{wellner2006classification,
  title = {{Classification of Discourse Coherence Relations: An Exploratory Study using Multiple Knowledge Sources}},
  author = {Wellner, Ben and Pustejovsky, James and Havasi, Catherine and Rumshisky, Anna and Saurí, Roser},
  booktitle = {Proceedings of the 7th SIGdial Workshop on Discourse and Dialogue},
  year = {2006},
  url = {https://aclanthology.org/W06-1317/}
}

References

19 references, 5 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. N. Asher and A. Lascarides. 2003. Logics of Conversation. Cambridge University Press, Cambridge, England. find
  2. J. Baldridge and A. Lascarides. 2005a. Annotating discourse structures for robust semantic interpretation. In Proceedings of the Sixth International Workshop on Computational Semantics, Tilburg, The Netherlands. find
  3. J. Baldridge and A. Lascarides. 2005b. Probabilistic head-driven parsing for discourse structure. In Proceedings of the Ninth Conference on Computational Natural Language Learning, Ann Arbor, USA. find
  4. T. Briscoe and J. Carroll. 2002. Robust accurate statistical annotation of general text. Proceedings of the Third International Conference on Language Resources and Evaluation (LREC 2002), Las Palmas, Canary Islands, May 2002, pages 1499–1504. find our card
  5. L. Carlson, D. Marcu, and M. E. Okurowski. 2003. Building a discourse-tagged corpus in the framework of rhetorical structure theory. In Janvan Kup-pelvelt and Ronnie Smith, editors, Current Directions in Discourse and Dialogue. Kluwer Academic Publishers. find our card
  6. K. Forbes, E. Miltsakaki, R. Prasad, A. Sakar, A. Joshi, and B. Webber. 2001. D-LTAG system: Discourse parsing with a lexicalized tree adjoining grammar. In Proceedings of the ESSLLI 2001: Workshop on Information Structure, Discourse Structure and Discourse Semantics. find
  7. J. Hobbs. 1985. On the coherence and structure of discourse. In CSLI Technical Report 85-37, Stanford, CA, USA. Center for the Study of Language and Information. find
  8. A. Killgarrif, P. Rychly, P. Smrz, and D. Tugwell. 2004. The sketch engine. In Proceedings of Euralex, Lorient, France, pages 105–116. find
  9. D. Lin. 1998. Automatic retrieval and clustering of similar words. In Proceedings of COLING-ACL, Montreal, Canada. find
  10. E. Miltsakaki, N. Dinesh, R. Prasad, A. Joshi, and B. Webber. 2005. Experiments on sense annotation and sense disambiguation of discourse connectives. In Proceedings of the Fourth Workshop on Treebanks and Linguistic Theories (TLT 2005), Barcelona, Catalonia. find
  11. J. Pustejovsky, C. Havasi, R. Saurí, P. Hanks, and A. Rumshisky. 2006. Towards a Generative Lexical resource: The Brandeis Semantic Ontology. In Language Resources and Evaluation Conference, LREC 2006, Genoa, Italy. find
  12. J. Pustejovsky. 1995. The Generative Lexicon. MIT Press, Cambridge, MA. find
  13. J. Pustejovsky. 2001. Type construction and the logic of concepts. In The Language of Word Meaning. Cambridge University Press. find
  14. R. Saurı́, M. Verhagen, and J. Pustejovsky. 2006. Annotating and recognizing event modality in text. In The 19th International FLAIRS Conference, FLAIRS 2006, Melbourne Beach, Florida, USA. find
  15. R. Soricut and D. Marcu. 2003. Sentence level discourse parsing using syntactic and lexical information. In Proceedings of the HLT/NAACL Conference, Edmonton, Canada. find our card
  16. M. Verhagen, I. Mani, R. Saurı́, R. Knippen, J. Littman, and J. Pustejovsky. 2005. Automating temporal annotation within TARSQI. In Proceedings of the ACL 2005. find
  17. B. Webber, A. Joshi, E. Miltsakaki, R. Prasad, N. Dinesh, A. Lee, and K. Forbes. 2005. A short introduction to the penn discourse TreeBank. In Copenhagen Working Papers in Language and Speech Processing. find our card
  18. B. Wellner and M. Vilain. 2006. Leveraging machine readable dictionaries in discriminative sequence models. In Language Resources and Evaluation Conference, LREC 2006, Genoa, Italy. find
  19. F. Wolf and E. Gibson. 2005. Representing discourse coherence: A corpus-based analysis. Computational Linguistics, 31(2):249–287. find our card

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/wellner-2006-classification-discourse-coherence-relations-exploratory/