Elamur Research⌕
☰
home / papers / 2017 / Apports des analyses syntaxiques pour la détection automatique de mentions dans un corpus de français oral

Apports des analyses syntaxiques pour la détection automatique de mentions dans un corpus de français oral

Summary

French coreference resolution still lacks a complete end-to-end pipeline. The earlier CROC system (Désoyer et al., 2015) groups mentions into chains, but it relies on gold mentions that are already annotated. This paper addresses the missing first step: locating mentions in raw text. It does so for transcribed spoken French, where tools designed for written language may perform worse. On the ANCOR corpus, the authors compare three strategies borrowed from coreference, anaphora and entity-tracking research:

  • pulling candidate constituents out of a full syntactic parse;
  • building mentions from shallow parsing (chunking), followed by post-processing;
  • casting mention-boundary detection as a learned sequence-labelling problem.

Quality is measured with precision, recall and F1, under both exact-span and head-based matching. None of the off-the-shelf parsers were adapted to speech. Even so, the resulting detection quality is roughly on par with leading mention detectors built for written text in other languages. The experiments therefore offer a baseline for a future end-to-end French coreference system, and the authors sketch ways to improve it.

Key points

  • Data: ANCOR (Muzerelle et al., 2013), a freely distributed corpus of spoken French annotated for coreference.
    • Singleton chains are retained.
    • Mentions can be nested.
    • Non-referential pronouns and noun phrases are excluded.
  • Three methods: deep-parse extraction, chunk-based extraction, and direct boundary labelling (sequence-labelling approaches such as MEMM, CRF and RNN). The methods were studied separately but could be combined.
  • Matching criteria: strict matching requires an identical span. Lenient matching requires the system mention to lie inside the gold mention and to contain its syntactic head (Recasens & Hovy, 2011). Recall is prioritised over precision.
  • Evaluation setup: the machine-learned methods were trained on 90% of ANCOR and tested on the remaining 10%. Comparisons with systems for other languages are only indicative.
  • Tools: the reference list cites the Talismane toolkit, a learned French chunker (Tellier et al., 2012) and large-scale CRF software (Lavergne et al., 2010). The exact scores fall in a part of the paper not included in the provided text.

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: Coreference resolution

Cited by

3 papers of our library cite it (of the 140673 whose bibliographies we read); 1 of them have a page here, 2 not yet.

Cite

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

@inproceedings{grobol2017apports,
  title = {{Apports des analyses syntaxiques pour la détection automatique de mentions dans un corpus de français oral}},
  author = {Grobol, Loïc and Tellier, Isabelle and Clergerie, Éric Villemonte de La and Dinarelli, Marco and Landragin, Frédéric},
  booktitle = {TALN 2017},
  year = {2017},
  url = {https://inria.hal.science/hal-01558711}
}

References

27 references, 2 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Abeillé A., Clément L. & Toussenel F. (2003). Building a Treebank for French, In A. Abeillé, Ed., Treebanks : Building and Using Parsed Corpora, p. 165–187. Springer Netherlands : Dordrecht. find
  2. Abney S. P. (1992). Parsing By Chunks, In R. C. Berwick, S. P. Abney & C. Tenny, Eds., Principle-Based Parsing : Computation and Psycholinguistics, p. 257–278. Springer Netherlands : Dordrecht. find
  3. Bove R. (2008). Automatic speech parsing : study of disfluencies. PhD thesis, Université de Provence -Aix-Marseille I. find
  4. Danlos L. (2005). ILIMP : Outil pour repérer les occurrences du pronom impersonnel il. In Actes de la 12ème conférence sur le Traitement Automatique des Langues Naturelles, p. 123–132, Dourdan, France : Association pour le Traitement Automatique des Langues. de Marneffe M.-C., Recasens M. & Potts C.… find
  5. Désoyer A., Landragin F., Tellier I., Lefeuvre A. & Antoine J.-Y. (2015). Coreference Resolution for Oral Corpus : a machine learning experiment with ANCOR corpus. Traitement Automatique des Langues, 55(2), 97–121. dos Santos C. N. & Carvalho D. L. (2011). Rule and tree ensembles for unrestricted co… find
  6. Finkel J. R. & Manning C. D. (2009). Nested named entity recognition. In Proceedings of the 2009 Conference on Empirical Methods in Natural Language Processing : Volume 1 - Volume 1, EMNLP ’09, p. 141–150, Stroudsburg, PA, USA : Association for Computational Linguistics. find
  7. Florian R., Pitrelli J. F., Roukos S. & Zitouni I. (2010). Improving mention detection robustness to noisy input. In Proceedings of the 2010 Conference on Empirical Methods in Natural Language Processing, EMNLP ’10, p. 335–345, Stroudsburg, PA, USA : Association for Computational Linguistics. find
  8. Jing H., Florian R., Luo X., Zhang T. & Ittycheriah A. (2003). HowtogetaChineseName(Entity) : Segmentation and combination issues. In Proceedings of the 2003 Conference on Empirical Methods in Natural Language Processing, EMNLP ’03, p. 200–207, Stroudsburg, PA, USA : Association for Computational Li… find
  9. Kummerfeld J. K., Bansal M., Burkett D. & Klein D. (2011). Mention detection : Heuristics for the ontonotes annotations. In Proceedings of the Fifteenth Conference on Computational Natural Language Learning : Shared Task, CONLL Shared Task ’11, p. 102–106, Stroudsburg, PA, USA : Association for Comp… find
  10. Lassalle E. (2015). Structured learning with latent trees : a joint approach to coreference resolution. PhD thesis. find
  11. Lavergne T., Cappé O. & Yvon F. (2010). Practical very large scale CRFs. In Proceedings the 48th Annual Meeting of the Association for Computational Linguistics (ACL), p. 504–513 : Association for Computational Linguistics. find
  12. Lee H., Chang A., Peirsman Y., Chambers N., Surdeanu M. & Jurafsky D. (2013). Deterministic coreference resolution based on entity-centric, precision-ranked rules. Computational Linguistics, 39(4), 885–916. find our card
  13. Longo L. (2013). Vers des moteurs de recherche ”intelligents” : un outil de détection automatique de thèmes. Méthode basée sur l’identification automatique des chaînes de référence. PhD thesis, Université de Strasbourg. find
  14. Muzerelle J., Lefeuvre A., Antoine J.-Y., Schang E., Maurel D., Villaneau J. & Eshkol I. (2013). Ancor, premier corpus de français parlé d’envergure annoté en coréférence et distribué librement. In Actes de la 20e conférence sur le Traitement Automatique des Langues Naturelles, p. 555–563, Les Sable… find
  15. Nguyen T. H., Sil A., Dinu G. & Florian R. (2016). Toward mention detection robustness with recurrent neural networks. CoRR, abs/1602.07749. — [arXiv:1602.07749](https://arxiv.org/abs/1602.07749) find
  16. Ogrodniczuk M., Wójcicka A., Głowińska K. & Kopeć M. (2014). Detection of Nested Mentions for Coreference Resolution in Polish, In A. Przepiórkowski & M. Ogrodniczuk, Eds., Advances in Natural Language Processing : 9th International Conference on NLP, PolTAL 2014, Warsaw, Poland, September 17-19, 20… find
  17. Poesio M., Stuckardt R. & Versley Y. (2016). Anaphora Resolution : Algorithms, Resources, and Applications. Theory and Applications of Natural Language Processing. Springer Berlin Heidelberg. find
  18. Qian D., Li W., Yuan C., Lu Q. & Wu M. (2007). Applying machine learning to Chinese entity detection and tracking. In Proceedings of the 8th International Conference on Computational Linguistics and Intelligent Text Processing, CICLing ’07, p. 154–165, Berlin, Heidelberg : Springer-Verlag. find
  19. Recasens M. (2010). Coreference : Theory, Annotation, Resolution and Evaluation. PhD thesis. find
  20. Recasens M. & Hovy E. (2011). Blanc : Implementing the Rand index for coreference evaluation. Nat. Lang. Eng., 17(4), 485–510. find
  21. Soraluze A., Arregi O., Arregi X., Ceberio K. & de Ilarraza A. D. (2012). Mention detection : First steps in the development of a Basque coreference resolution system. In J. Jancsary, Ed., Proceedings of KONVENS 2012, p. 128–136 : ÖGAI. find
  22. Tellier I., Duchier D., Eshkol I., Courmet A. & Martinet M. (2012). Apprentissage automatique d’un chunker pour le français. In Actes de la 19e conférence sur le Traitement Automatique des Langues Naturelles, p. 431–438, Grenoble, France : Association pour le Traitement Automatique des Langues. find
  23. Tellier I., Eshkol-Taravella I., Dupont Y. & Wang I. (2014). Peut-on bien chunker avec de mauvaises étiquettes POS ? In Actes de la 21e conférence sur le Traitement Automatique des Langues Naturelles, p. 125–136, Marseille, France : Association pour le Traitement Automatique des Langues. find
  24. Trouilleux F. (2001). Referential links identification and automatic interpretation of pronominal expressions in French texts. PhD thesis, Université Blaise-Pascal, Clermont-Ferrand. find
  25. Urieli A. (2013). Robust French syntax analysis : reconciling statistical methods and linguistic knowledge in the Talismane toolkit. PhD thesis, Université de Toulouse II le Mirail. find
  26. Uryupina O. & Moschitti A. (2013). Multilingual mention detection for coreference resolution. In IJCNLP, p. 100–108. find
  27. Versley Y., Ponzetto S. P., Poesio M., Eidelman V., Jern A., Smith J., Yang X. & Moschitti A. (2008). BART : A modular toolkit for coreference resolution. In Proceedings of the 46th Annual Meeting of the Association for Computational Linguistics on Human Language Technologies : Demo Session, HLT-Dem… find our card

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/papers/grobol-2017-apports-analyses-syntaxiques-detection/