Predicting Conjunct Propagation and Other Extended Stanford Dependencies

What it is about

The authors add the non-basic layer of Stanford Dependencies (conjunct propagation, external subjects of open complements and the function of relativizers) on top of an ordinary dependency tree. The method is hybrid: a machine learner predicts the extra arcs from the base tree and the morphology, with rules for the rest. The training and test data is the Turku Dependency Treebank for Finnish, 15,126 sentences, which already has both layers. The method reaches 93.1% F1 on the extra dependencies. The authors also convert between the conjunction-as-head style and the first-conjunct style of Stanford Dependencies, which loses only a few structures: forward and backward conversion give 92.6% precision and 92.3% recall on propagated dependencies.

Ideas

• I1. en/src/ctx.rs and en/src/parse.rs: a second, learned pass that predicts propagated conj arcs (one decision per governor and conjunct pair), and measure F1 on EWT test against the enhanced layer, treated as a derived layer.
• I2. prag (world): propagate a subject or object shared by coordinated verbs to every conjunct in the frames, and measure event-argument recall on 50 EWT sentences with coordinated verbs by hand count.
• I3. treeagree and annot: round-trip the silver trees through the conjunction-as-head style and back, count the arcs lost, and measure the round-trip recall on EWT train against the paper's 92.3%.
• I4. en/src/explain.rs: for each coordination the transition parser gets wrong, print whether the first conjunct, the conjunction or a chained conjunct was the head; count the cases per error type on EWT test.

For Mova

Touches 24. Coordination and subordination and 20. Syntax: the Universal Dependencies frame. The propagated arcs matter for the prag and world layers, which read UD trees into frames. Measure first the share of EWT test coordinations where a shared dependent is missing from the frames.

Cite

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

@inproceedings{nyblom2013predicting,
  title = {{Predicting Conjunct Propagation and Other Extended Stanford Dependencies}},
  author = {Nyblom, Jenna and Kohonen, Samuel and Haverinen, Katri and Salakoski, Tapio and Ginter, Filip},
  booktitle = {Proceedings of the Second International Conference on Dependency Linguistics (DepLing 2013)},
  year = {2013},
  url = {https://aclanthology.org/W13-3728/}
}

References

28 references, 4 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Rajeev Agarwal and Lois Boggess. 1992. A simple but useful approach to conjunct identification. In Proceedings of ACL’92, pages 15–21. find
2. Bernd Bohnet. 2010. Top accuracy and fast dependency parsing is not a contradiction. In Proceedings of COLING’10, pages 89–97. find
3. Joan Bresnan. 2001. Lexical-functional syntax, volume 16 of Blackwell Textbooks in Linguistics. Blackwell. find
4. Richard Campbell. 2004. Using linguistic principles to recover empty categories. In Proceedings of ACL, pages 646–653. find
5. Francis Chantree, Adam Kilgarriff, Anne de Roeck, and Alistair Willis. 2005. Disambiguating coordinations using word distribution information. In Proceedings of RANLP’05. find
6. Andrew B. Clegg and Adrian Shepherd. 2007. Benchmarking natural-language parsers for biological applications using dependency graphs. BMC Bioinformatics, 8(1):24. find
7. Miriam Goldberg. 1999. An unsupervised model for statistically determining coordinate phrase attachment. In Proceedings of ACL’99, pages 610–614. find
8. Jan Hajič et al. 2009. The conll-2009 shared task: syntactic and semantic dependencies in multiple languages. In Proceedings of CoNLL’09, pages 1–18. find our card
9. Péter Halácsy, András Kornai, and Csaba Oravecz. 2007. HunPos – an open source trigram tagger. In Proceedings of ACL’07, Companion Volume, pages 209–212. find
10. Katri Haverinen, Filip Ginter, Veronika Laippala, Timo Viljanen, and Tapio Salakoski. 2010. Dependency-based propbanking of clinical Finnish. In Proceedings of LAW IV, pages 137–141. find
11. Katri Haverinen, Filip Ginter, Veronika Laippala, Samuel Kohonen, Timo Viljanen, Jenna Nyblom, and Tapio Salakoski. 2011. A dependency-based analysis of treebank annotation errors. In Proceedings of Depling’11, pages 115–124. find
12. Katri Haverinen. 2012. Syntax annotation guidelines for the Turku Dependency Treebank. Technical Report 1034, Turku Centre for Computer Science, January. find
13. Thorsten Joachims. 1999. Making large-scale SVM learning practical. In Advances in Kernel Methods - Support Vector Learning, pages 169–184. MIT Press. find
14. Daisuke Kawahara and Sadao Kurohashi. 2007. Probabilistic coordination disambiguation in a fully-lexicalized Japanese parser. In Proceedings of EMNLP-CoNLL’07, pages 306–314. find
15. Daisuke Kawahara and Sadao Kurohashi. 2011. Generative modeling of coordination by factoring parallelism and selectional preferences. In Proceedings of IJCNLP’11, pages 456–464. find
16. Jin-Dong Kim, Sampo Pyysalo, Tomoko Ohta, Robert Bossy, Ngan Nguyen, and Jun’ichi Tsujii. 2011. Overview of bionlp shared task 2011. In Proceedings of BioNLP Shared Task 2011 Workshop, pages 1–6. Association for Computational Linguistics. find
17. Tracy Holloway King, Crouch Richard, Stefan Riezler, Mary Dalrymple, and Ronald M. Kaplan. 2003. The PARC 700 dependency bank. In Proceedings of LINC’03, pages 1–8. find
18. Dekang Lin. 1998. A dependency-based method for evaluating broad-coverage parsers. Natural Language Engineering, 4(2):97–114. find
19. Marie-Catherine de Marneffe and Christopher Manning. 2008. Stanford typed dependencies manual. Technical report, Stanford University, September. find
20. Ryan McDonald and Fernando Pereira. 2006. Online learning of approximate dependency parsing algorithms. In Proceedings of the 11th Conference of the European Chapter of the Association for Computational Linguistics (EACL), pages 81–88. find our card
21. Joakim Nivre, Johan Hall, Jens Nilsson, Atanas Chanev, Gülşen Eryiǧit, Sandra Kübler, Svetoslav Marinov, and Erwin Marsi. 2007. MaltParser: A language-independent system for data-driven dependency parsing. Natural Language Engineering, 13(2):95–135. find
22. Martha Palmer, Daniel Gildea, and Paul Kingsbury. 2005. The Proposition Bank: An annotated corpus of semantic roles. Computational Linguistics, 31(1):71–106. ↗ our card
23. Philip Resnik. 1999. Semantic similarity in a taxonomy: An information-based measure and its application to problems of ambiguity in natural language. Journal of Artificial Intelligence Research, 11:95– 130. find
24. Kenji Sagae and Jun’ichi Tsujii. 2008. Shift-reduce dependency dag parsing. In Proceedings of the 22nd International Conference on Computational Linguistics (Coling 2008) - Volume 1, pages 753– 760. find
25. Roy Schwartz, Omri Abend, and Ari Rappoport. 2012. Learnability-based syntactic annotation design. In Proceedings of the 24th International Conference on Computational Linguistics (Coling 2012), pages 2405–2422. find
26. Daniel Sleator and Davy Temperley. 1993. Parsing English with a Link Grammar. In Proceedings of IWPT’93, pages 277–291. find our card
27. Gertjan van Noord. 2007. Using-self-trained bilexical preferences to improve disambiguation accuracy. In Proceedings of IWPT’07, pages 1–10. find
28. Deniz Yuret, Laura Rimmell, and Aydin Han. 2012. Parser evaluation using textual entailments. Language Resources and Evaluation, pages 1–21. find

This page in Markdown → https://md.mova.elamur.ai/free/en/docs/papers/bank-2013-nyblom-predicting-conjunct-propagation-other-extended/