Elamur Research⌕
☰
home / papers / 2012 / Event Nominals: Annotation Guidelines and a Manually Annotated Corpus in French

Event Nominals: Annotation Guidelines and a Manually Annotated Corpus in French

Summary

The authors work on detecting events in texts for information extraction tasks. They focus on nouns and noun phrases that denote events, because such resources are far less developed than verbal ones. At the time, the only resource for French was FR-TimeBank, annotated according to TimeML, where nouns are a secondary concern. The authors treat an event as a named entity, in line with the guidelines of the Quaero project. They propose a typology of events by modality, frequency and temporal anchoring. They also give rules for telling an eventive reading from a non-eventive one and for setting the boundaries of an annotation. Following these guidelines, they manually annotated a corpus of French news and studied how event-denoting words behave. The analysis showed that such nouns are highly ambiguous. Common assumptions about plurals and definite articles were not confirmed. Experiments with lexicons showed that lexicons alone are not enough for reliable extraction. The work provides a corpus for evaluating systems that automatically detect event nouns in French.

Key points

  • Corpus. 192 news texts, 47 646 words, 1844 annotated events. The Le Monde part contains 83 texts and 1107 events. Another 109 texts are shared with FR-TimeBank and contain 737 events. The corpus is available on request.
  • Agreement. Between the two annotators, Cohen's Kappa = 0.808, computed only on the heads of event phrases. Against the FR-TimeBank annotation, Kappa = 0.704. The main differences are that the authors do not annotate states, but they do annotate named events such as the names of wars.
  • Typology. Modality can be factual, hypothetical, non-factual or abstract. By frequency, an event is single or recurring. In time, it lies before, during or after the moment of speech. Metonymy is described separately, where place names and dates denote events (Chornobyl, September 11).
  • Ambiguity. Among nouns that occur in the corpus more than once, only 31% denote an event every time. Event nouns occur with an indefinite article more often than other nouns (14.3% vs. 6.2%).
  • Lexicons. Even an "ideal" lexicon of unambiguous event nouns finds only 42% of events (F = 0.593). When the lexicon was built on 75% of the corpus, F on the rest of the text dropped to 0.395 for the unambiguous lexicon and to 0.415 for the full one.

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: Treebanks

Cite

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

@inproceedings{arnulphy2012event,
  title = {{Event Nominals: Annotation Guidelines and a Manually Annotated Corpus in French}},
  author = {Arnulphy, Béatrice and Tannier, Xavier and Vilnat, Anne},
  booktitle = {Proceedings of the Eighth International Conference on Language Resources and Evaluation (LREC'12)},
  year = {2012},
  doi = {10.63317/2w3b8pv2g6y6},
  url = {https://doi.org/10.63317/2w3b8pv2g6y6}
}

References

11 references, 0 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. André Bittar. 2010. Building a TimeBank for French: A Reference Corpus Annotated According to the ISO-TimeML Standard. Ph.D. thesis, Université Paris Diderot. find
  2. Laura Calabrese Steimberg. 2008. Les héméronymes. Ces évènements qui font date, ces dates qui deviennent évènements. Mots. Les langages du politique, 3:115– 128. find
  3. Jacob Cohen. 1960. A Coefficient of Agreement for Nominal Scales. Educational and Psychological Measurement, 20(1):37–46. find
  4. Cassandre Creswell, Matthew J. Beal, John Chen, Thomas L. Cornell, Lars Nilsson, and Rohini K. Srihari. 2006. Automatically Extracting Nominal Mentions of Events with a Bootstrapped Probabilistic Classifier. In Proceedings of the COLING/ACL 2006 Main Conference, pages 168–175, Sydney, Australia, jul… find
  5. Maud Ehrmann and Caroline Hagège. 2009. Proposition de caractérisation et de typage des expressions temporelles en contexte. In 16ème, Senlis, France, June. ATALA. find
  6. J R Landis and G G Koch. 1977. The Measurement of Observer Agreement for Categorical Data. Biometrics, 33. find
  7. Michelle Lecolle. 2009. Éléments pour la caractérisation des toponymes en emploi événementiel. In Ivan Evrard, Michel Pierrard, Laurence Rosier, and Dan Van Raem-donck, editors, Les sens en marge - Représentations linguistiques et observables discursifs : actes du colloque international de Bruxelles… find
  8. James Pustejovsky, José Castaño, Robert Ingria, Roser Saurí, Robert Gaizauskas, Andrea Setzer, and Graham Katz. 2003. TimeML: Robust Specification of Event and Temporal Expressions in Text. In IWCS-5, Fifth International Workshop on Computational Semantics. find
  9. Sophie Rosset, Cyril Grouin, and Pierre Zweigenbaum, 2011. Entités Nommées Structurées : guide d’annotation Quaero. LIMSI–CNRS, Orsay, France. http://www.quaero.org/media/files/bibliographie/ quaero-guide-annotation-2011.pdf. ↗
  10. Irene Russo, Tommaso Caselli, and Francesco Rubino. 2011. Recognizing deverbal events in context. In Proceedings of 12th International Conference on Intelligent Text Processing and Computational Linguistics (CI-CLing 2011), poster session, Tokyo, Japan, feb. Springer. find
  11. Xavier Tannier, Véronique Moriceau, Béatrice Arnulphy, and Ruixin He. 2012. Evolution of Event Designation in Media: Preliminary Study. In Proceedings of the Eighth International Language Resources and Evaluation (LREC’2012), Istanbul, Turkey, May. find

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/papers/arnulphy-2012-event-nominals-annotation-guidelines-manually/