Elamur Research⌕
☰
home / papers / 2010 / Statistical French Dependency Parsing: Treebank Conversion and First Results

Statistical French Dependency Parsing: Treebank Conversion and First Results

Summary

The paper addresses the lack of training data for statistical dependency parsing of French. The authors automatically converted the French Treebank (FTB), a corpus of constituency trees built from the newspaper Le Monde, into typed projective dependency trees in CoNLL format. First, the corpus was preprocessed: some syntactically regular compounds were split into separate words, and prepositions and complementizers were made heads. Heads were then determined using head-propagation rules. Grammatical functions missing from the original annotation were added by heuristics. The quality of the conversion was evaluated on a manually corrected sample. This also showed how many errors result from the projectivity requirement, especially for non-local dependencies such as extractions. The authors then compared two approaches: the MST parser trained directly on dependencies, and the Berkeley parser followed by function labeling and conversion of its output into dependencies. MST achieves a slightly higher UAS and a noticeably higher LAS. The work produced the first usable dependency resource for French. It also provided a common basis for comparing constituency and dependency parsers.

Key points

  • The FTB contains 12,531 sentences. In the new version, FTB-UC, 3,072 of the 6,125 distinct compounds were split. This yields 350,951 tokens, compared with 339,522 when all compounds are merged.
  • Heads were determined with the Magerman method using the improved rules of Arun & Keller. Auxiliary verbs depend on the participle or infinitive, and the first conjunct becomes the head of a coordination. For prepositional and subordinate dependents of non-verbal heads, the underspecified label dep was kept.
  • The gold sample FTB-120 consists of the first 120 sentences of the corpus (about 3000 tokens), corrected manually. The authors described typical non-projective cases: extraction from a clause, the pronoun dont, the clitic en, and comparative constructions.
  • With an accurate POS tagger, MST reached a UAS of 90.3% and a LAS of 87.6%. Berkeley with function labeling (BKY+FLabeler) reached a UAS of 89.6% and a LAS of 85.6%. The difference is statistically significant: p < 0.01 by the chi-square test.
  • For comparison with CoNLL 2007, UAS including punctuation is reported: 87.52% for the best MST configuration versus 86.85% for BKY+FLabeler.

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: Dependency parsing · Treebanks

Institutions

🔬Institut national de recherche en sciences et technologies du numériqueLe Chesnay-Rocquencourt · France55 works · 16 624 citations🎓Université Paris DiderotParis · France13 works · 1 973 citations🔬Analyse Linguistique Profonde à Grande EchelleParis · France11 works · 1 087 citations

Cite

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

@inproceedings{candito2010statistical,
  title = {{Statistical French Dependency Parsing: Treebank Conversion and First Results}},
  author = {Candito, Marie and Crabbé, Benoît and Denis, Pascal},
  booktitle = {Proceedings of the Seventh International Conference on Language Resources and Evaluation (LREC'10)},
  year = {2010},
  doi = {10.63317/3kefxppz8men},
  url = {https://doi.org/10.63317/3kefxppz8men}
}

References

29 references, 14 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Anne Abeillé and Nicolas Barrier. 2004. Enriching a french treebank. In Proc. of LREC’04, Lisbon, Portugal. find
  2. Abhishek Arun and Frank Keller. 2005. Lexicalization in crosslinguistic probabilistic parsing: The case of french. In Proceedings of the 43rd Annual Meeting of the Association for Computational Linguistics, pages 306–313, Ann Arbor, MI. find
  3. Marie Candito and Benoît Crabbé. 2009. Improving generative statistical parsing with semi-supervised word clustering. In Proceedings of the 11th International Conference on Parsing Technologies (IWPT’09), pages 138–141, Paris, France, October. Association for Computational Linguistics. find
  4. Marie Candito, Benoit Crabbé, Pascal Denis, and François Guérin. 2009. Analyse syntaxique du français : des constituants aux dépendances. In Proc. of TALN’09, Senlis, France. find
  5. Eugene Charniak. 2000. A maximum-entropy-inspired parser. In Proceedings of the 1st Annual Meeting of the North American Chapter of the ACL (NAACL), Seattle. find our card
  6. Y. J. Chu and T. H. Liu. 1965. On the shortest arborescence of a directed graph. Science Sinica, 14:1396– 1400. find
  7. Michael Collins. 1999. Head Driven Statistical Models for Natural Language Parsing. Ph.D. thesis, University of Pennsylvania, Philadelphia. find our card
  8. Michael Collins. 2002. Discriminative training methods for hidden markov models: Theory and experiments with perceptron algorithms. In EMNLP-2002. find our card
  9. Benoit Crabbé and Marie Candito. 2008. Expériences d’analyse syntaxique statistique du français. In Proc. of TALN’08, pages 45–54, Senlis, France. find
  10. Koby Crammer and Yoram Singer. 2001. Ultraconservative online algorithms for multiclass problems. Journal of Machine Learning Research, 3. find
  11. Marie-Catherine de Marneffe, Bill MacCartney, and Christopher D. Manning. 2006. Generating typed dependency parses from phrase structure parses. In Proc. of LREC’06, Genova, Italy. find our card
  12. Pascal Denis and Benoît Sagot. 2009. Coupling an annotated corpus and a morphosyntactic lexicon for state-of-the-art pos tagging with less human effort. In Proc. of PACLIC, Hong Kong, China. find
  13. Jack R. Edmonds. 1967. Optimum branchings. Journal of Research of the National Bureau of Standards, 71B:233–240. find
  14. Jason Eisner. 1996. Three new probabilistic models for dependency parsing: An exploration. In Proceedings of the 16th International Conference on Computational Linguistics (COLING-96), pages 340–345, Copenhagen, August. find our card
  15. Richard Johansson and Pierre Nugues. 2007. Extended constituent-to-dependency conversion for English. In Proceedings of NODALIDA 2007, Tartu, Estonia. find our card
  16. Roger Levy and Galen Andrew. 2006. Tregex and tsurgeon: tools for querying and manipulating tree data structures. In Proc. of LREC’06, Geneva, Italy. find
  17. D.M. Magerman. 1995. Statistical decision-tree models for parsing. In Proc. of ACL’95, pages 276–283, Morristown, NJ, USA. find our card
  18. Takuya Matsuzaki, Yusuke Miyao, and Jun’ichi Tsujii. 2005. Probabilistic cfg with latent annotations. In Proceedings of the 43rd Annual Meeting of the Association for Computational Linguistics (ACL), pages 75–82. find our card
  19. Ryan T. McDonald and Fernando C. N. Pereira. 2006. Online learning of approximate dependency parsing algorithms. In Proc. of EACL’06. find our card
  20. Ryan T. McDonald, Koby Crammer, and Fernando C. N. Pereira. 2005. Online large-margin training of dependency parsers. In Proc. of ACL’05, Ann Arbor, USA. find our card
  21. Ryan Mcdonald, Kevin Lerman, and Fernando Pereira. 2006. Multilingual dependency analysis with a two-stage discriminative parser. In Proc. of CoNLL’06, New York, USA. find our card
  22. Alexis Nasr. 2006. Grammaires de dépendances génératives probabilistes. modèle théorique et application à un corpus arboré du français. Traitement Automatique des Langues, 46(1). find
  23. Joakim Nivre, Johan Hall, Sandra Kübler, Ryan Mc-Donald, Jens Nilsson, Sebastian Riedel, and Deniz Yuret. 2007. find
  24. The CoNLL 2007 shared task on dependency parsing. In Proceedings of the CoNLL Shared Task Session of EMNLP-CoNLL 2007, pages 915–932, Prague, Czech Republic, June. Association for Computational Linguistics. find our card
  25. Patrick Paroubek, Louis-Gabriel Pouillot, Isabelle Robba, and Anne Vilnat. 2005. Easy : campagne d’évaluation des analyseurs syntaxiques. In Proceedings of TALN’05, EASy workshop : campagne d’évaluation des analyseurs syntaxiques, Dourdan. find
  26. Slav Petrov, Leon Barrett, Romain Thibaux, and Dan Klein. 2006. Learning accurate, compact, and interpretable tree annotation. In Proceedings of the 21st International Conference on Computational Linguistics and 44th Annual Meeting of the Association for Computational Linguistics, Sydney, Australia, July. Association for Computational Linguistics. find our card
  27. Natalie Schluter and Josef van Genabith. 2007. Preparing, restructuring, and augmenting a french treebank: Lexicalised parsers or coherent treebanks? In Proc. of PACLING 07, Melbourne, Australia. find
  28. Djamé Seddah, Marie Candito, and Benoit Crabbé. 2009. Cross parser evaluation and tagset variation: a french treebank study. In Proceedings of the 11th Internation Conference on Parsing Technologies (IWPT’09), pages 150–161, Paris, France, October. ↗
  29. H. Yamada and Y. Matsumoto. 2003. Statistical dependency analysis with support vector machines. In The 8th International Workshop of Parsing Technologies (IWPT2003). find our card

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/candito-2010-statistical-french-dependency-parsing-treebank/