Elamur Research⌕
☰
home / papers / 2021 / End-to-end Neural Information Status Classification

End-to-end Neural Information Status Classification

Paper · 2021 · Yufang Hou

Summary

Earlier work on information status (IS) classification, which labels each discourse entity as old, mediated (with several subtypes such as bridging) or new, assumed that gold mentions or syntactic parse trees were available. This paper builds a fully end-to-end neural pipeline that starts from raw text. It has two transformer-based models. MenExt scores every span as mention or non-mention, keeping singletons as well as coreferent mentions. ISAssign then gives each predicted mention one of eight fine-grained IS labels. Both models mark the target span with special boundary tokens and classify it from the encoder states of those markers. ISAssign also receives a cue saying whether the same string was mentioned earlier. On ISNotes, ISAssign sets a new state of the art for fine-grained IS classification with gold mentions, and the full system clearly beats baselines in the end-to-end setting. Trained only on ISNotes news, it also recognises referential bridging anaphors on BASHI and SciCorp about as well as a syntax-based system trained in-domain. This shows that mentions can be found reliably without parsing.

Key points

  • The scheme has eight IS categories from Markert et al. (2012): old; m/syntactic, m/worldKnow, m/bridging, m/comparative, m/aggregate and m/function; and new. The main benchmark is ISNotes.
  • Mention extraction is trained on all spans of up to L = 10 words, following Lee et al. (2017). At inference it scores all O(N²) spans with no pruning.
  • In ISNotes, about 10% of mentions are longer than ten words, and roughly 80% of those are singletons. Even so, the model reaches an F-score of 83.4 on this long-mention group.
  • On bridging recognition, the ISNotes-trained system gets higher recall than Yu and Poesio (2020): 59.7 vs 34.2 on BASHI and 47.6 vs 35.7 on SciCorp. Its F is comparable, at 35.8 vs 34.3 and 41.0 vs 39.8.
  • The dominant errors are confusions between bridging and new. 202 bridging anaphors were labelled new, and 142 new mentions were labelled bridging, often generic relational nouns such as "revenue" or "costs".

Interesting ideas

  • I1. Marking a candidate span with two inserted boundary tokens and classifying from their contextual states gives one generic span classifier that works for any span, with no parse or width features.
  • I2. Train on short spans but score every span at inference without pruning. Dense mention/non-mention signal over short spans was enough to generalise to longer ones.
  • I3. A cheap discourse feature, whether the same string occurred earlier in the text, strongly helps detect "old" entities.
  • I4. Treating mediated entities as subtypes (inferable through bridging, world knowledge, syntactic linkage, comparison, aggregation or function values) separates how an entity becomes accessible from mere coreference.
  • I5. The error analysis shows that relational nouns used generically are the main trap for bridging detection.

For Mova

How these ideas could serve Mova, our small open language model — its files, stages and tests:

  • I1 (frames): A marker-pair classifier could act as an independent witness that checks the parser's noun-phrase boundaries before mentions enter coreference and frame filling.
  • I2 (learn): Silver training sets can be capped at manageable span lengths while evaluation still covers all spans, with long-span performance tracked separately as a check.
  • I3 (world): The text-world builder could record string-repeat flags per entity as an explainable first signal for given versus new before full coreference decides.
  • I4 (world): When an entity first appears, Mova could log an accessibility label, e.g. "bridging from the murder event", so answers can explain why a definite noun phrase needs no prior mention.
  • I5 (frames): The bridging resolver could check whether a relational noun's implicit argument is actually filled in context, using a WordNet/FrameNet relational-noun list, before positing an anaphoric link.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Institutions

🔬IBM Research - IrelandDublin · Ireland8 works · 889 citations

📚 Shelves 4 Computer Science · 9,086 works Artificial Intelligence · 8,889 works Arts and Humanities · 9,296 works Language and Linguistics · 9,206 works

Cite

Cite as Yufang Hou. 2021. End-to-end Neural Information Status Classification. Findings of the Association for Computational Linguistics: EMNLP 2021.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2021.findings-emnlp.119

@inproceedings{hou2021end,
  title = {{End-to-end Neural Information Status Classification}},
  author = {Yufang Hou},
  booktitle = {Findings of the Association for Computational Linguistics: EMNLP 2021},
  year = {2021},
  doi = {10.18653/v1/2021.findings-emnlp.119},
  url = {https://doi.org/10.18653/v1/2021.findings-emnlp.119}
}

References

36 references, 25 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Aoife Cahill and Arndt Riester. 2012. Automatically acquiring fine-grained information status distinctions in German. In Proceedings of the SIGdial 2012 Conference: The 13th Annual Meeting of the Special Interest Group on Discourse and Dialogue, Seoul, Korea, 5–6 July 2012, pages 232–236. find
  2. Herbert H. Clark. 1975. Bridging. In Proceedings of the Conference on Theoretical Issues in Natural Language Processing, Cambridge, Mass., June 1975, pages 169–174. find
  3. Kevin Clark and Christopher D. Manning. 2016. Improving coreference resolution by learning entity-level distributed representations. In Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 643–653, Berlin, Germany. Association for Computational Linguistics. ↗ our card
  4. M. A. K. Halliday. 1967. Notes on transitivity and theme in English, Part 2. Journal of Linguistics, 3:199–244. find
  5. Yufang Hou. 2016. Incremental fine-grained information status classification using attention-based LSTMs. In Proceedings of the 26th International Conference on Computational Linguistics, Osaka, Japan, 11–16 December 2016, pages 1880–1890. our card
  6. Yufang Hou. 2018. A deterministic algorithm for bridging anaphora resolution. In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, Brussels, Belgium, 31 October– 4 November 2018, pages 1938–1948. our card
  7. Yufang Hou. 2020a. Bridging anaphora resolution as question answering. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, Seattle, Wash., 5–10 July 2020, pages 1428–1438. our card
  8. Yufang Hou. 2020b. Fine-grained information status classification using discourse context-aware BERT. In Proceedings of the 28th International Conference on Computational Linguistics, pages 6101–6112, Barcelona, Spain (Online). International Committee on Computational Linguistics. our card
  9. Yufang Hou, Katja Markert, and Michael Strube. 2013a. Cascading collective classification for bridging anaphora recognition using a rich linguistic feature set. In Proceedings of the 2013 Conference on Empirical Methods in Natural Language Processing, Seattle, Wash., 18–21 October 2013, pages 814– 820. ↗ our card
  10. Yufang Hou, Katja Markert, and Michael Strube. 2013b. Global inference for bridging anaphora resolution. In Proceedings of the 2013 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Atlanta, Georgia, 9–14 June 2013, pages 907–917. our card
  11. Yufang Hou, Katja Markert, and Michael Strube. 2018. Unrestricted bridging resolution. Computational Linguistics, 44(2):237–284. ↗ our card
  12. Nikolaos Kolitsas, Octavian-Eugen Ganea, and Thomas Hofmann. 2018. End-to-end neural entity linking. In Proceedings of the 22nd Conference on Computational Natural Language Learning, pages 519–529, Brussels, Belgium. Association for Computational Linguistics. ↗
  13. Heeyoung Lee, Yves Peirsman, Angel Chang, Nathanael Chambers, Mihai Surdeanu, and Dan Jurafsky. 2011. Stanford’s multi-pass sieve coreference resolution system at the CoNLL-2011 shared task. In Proceedings of the Shared Task of the 15th Conference on Computational Natural Language Learning, Portland, Oreg., 23–24 June 2011, pages 28–34. our card
  14. Malvina Nissim, Shipara Dingare, Jean Carletta, and Mark Steedman. 2004. An annotation scheme for information status in dialogue. In Proceedings of the 4th International Conference on Language Resources and Evaluation, Lisbon, Portugal, 26–28 May 2004, pages 1023–1026. ↗
  15. Kenton Lee, Luheng He, Mike Lewis, and Luke Zettlemoyer. 2017. End-to-end neural coreference resolution. In Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing, pages 188–197, Copenhagen, Denmark. Association for Computational Linguistics. ↗ our card
  16. Massimo Poesio, Yulia Grishina, Varada Kolhatkar, Nafise Sadat Moosavi, Ina Rösiger, Adam Roussel, Fabian Simonjetz, Alexandra Uma, Olga Uryupina, Juntao Yu, and Heike Zinsmeister. 2018. Anaphora resolution with the ARRAU corpus. In Proceedings of the Workshop on Computational Models of Reference, Anaphora and Coreference. New Orleans, Louisiana, June 6, 2018, pages 11–22. ↗ our card
  17. Yinhan Liu, Myle Ott, Naman Goyal, Jingfei Du, Mandar Joshi, Danqi Chen, Omer Levy, Mike Lewis, Luke Zettlemoyer, and Veselin Stoyanov. 2019. Roberta: A robustly optimized bert pretraining approach. find
  18. Massimo Poesio, Rahul Mehta, Axel Maroudas, and Janet Hitzeman. 2004. Learning to resolve bridging references. In Proceedings of the 42nd Annual Meeting of the Association for Computational Linguistics, Barcelona, Spain, 21–26 July 2004, pages 143–150. ↗ our card
  19. Sebastian Löbner. 1985. Definites. Journal of Semantics, 4:279–326. find
  20. Ellen F. Prince. 1981. Towards a taxonomy of given-new information. In P. Cole, editor, Radical Pragmatics, pages 223–255. Academic Press, New York, N.Y. find
  21. Christopher Manning, Mihai Surdeanu, John Bauer, Jenny Finkel, Steven Bethard, and David McClosky. 2014. The Stanford CoreNLP natural language processing toolkit. In Proceedings of 52nd Annual Meeting of the Association for Computational Linguistics: System Demonstrations, pages 55–60, Baltimore, Maryland. Association for Computational Linguistics. ↗ our card
  22. Katja Markert, Yufang Hou, and Michael Strube. 2012. Collective classification for fine-grained information status. In Proceedings of the 50th Annual Meeting of the Association for Computational Linguistics, Jeju Island, Korea, 8–14 July 2012, pages 795–804. our card
  23. Mike Mintz, Steven Bills, Rion Snow, and Daniel Jurafsky. 2009. Distant supervision for relation extraction without labeled data. In Proceedings of the Joint Conference of the 47th Annual Meeting of the ACL and the 4th International Joint Conference on Natural Language Processing of the AFNLP, pages 1003–1011, Suntec, Singapore. Association for Computational Linguistics. find
  24. Natalia N. Modjeska, Katja Markert, and Malvina Nissim. 2003. Using the web in machine learning for other-anaphora resolution. In Proceedings of the 2003 Conference on Empirical Methods in Natural Language Processing, pages 176–183. our card
  25. Marius Mosbach, Maksym Andriushchenko, and Dietrich Klakow. 2020. On the stability of fine-tuning bert: Misconceptions, explanations, and strong baselines. ↗
  26. Vincent Ng. 2010. Supervised noun phrase coreference research: The first fifteen years. In Proceedings of the 48th Annual Meeting of the Association for Computational Linguistics, pages 1396–1411, Uppsala, Sweden. Association for Computational Linguistics. our card
  27. Ellen F. Prince. 1992. The ZPG letter: Subjects, definiteness, and information-status. In W.C. Mann and S.A. Thompson, editors, Discourse Description. Diverse Linguistic Analyses of a Fund-Raising Text, pages 295–325. John Benjamins, Amsterdam. ↗
  28. Altaf Rahman and Vincent Ng. 2012. Learning the fine-grained information status of discourse entities. In Proceedings of the 13th Conference of the European Chapter of the Association for Computational Linguistics, Avignon, France, 23–27 April 2012, pages 798–807. our card
  29. Ina Rösiger. 2016. SciCorp: A corpus of English scientific articles annotated for information status analysis. In Proceedings of the Tenth International Conference on Language Resources and Evaluation (LREC’16), pages 1743–1749, Portorož, Slovenia. European Language Resources Association (ELRA). our card
  30. Ina Rösiger. 2018. BASHI: A corpus of wall street journal articles annotated with bridging links. In Proceedings of the 11th International Conference on Language Resources and Evaluation, Miyazaki, Japan, 7–12 May 2018, pages 382–388. ↗ our card
  31. Ina Rösiger, Arndt Riester, and Jonas Kuhn. 2018. Bridging resolution: Task definition, corpus resources and rule-based experiments. In Proceedings of the 27th International Conference on Computational Linguistics, Santa Fe, New-Mexico, USA, 20– 26 August 2018, pages 3516–3528. our card
  32. Sam Wiseman, Alexander M. Rush, Stuart Shieber, and Jason Weston. 2015. Learning anaphoricity and antecedent ranking features for coreference resolution. In Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing (Volume 1: Long Papers), pages 1416–1426, Beijing, China. Association for Computational Linguistics. ↗ our card
  33. Thomas Wolf, Lysandre Debut, Victor Sanh, Julien Chaumond, Clement Delangue, Anthony Moi, Pier-ric Cistac, Tim Rault, Remi Louf, Morgan Funtowicz, Joe Davison, Sam Shleifer, Patrick von Platen, Clara Ma, Yacine Jernite, Julien Plu, Canwen Xu, Teven Le Scao, Sylvain Gugger, Mariama Drame, Quentin Lhoest, and Alexander Rush. 2020. Transformers: State-of-the-art natural language processing. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing: System Demonstrations, pages 38–45, Online. Association for Computational Linguistics. ↗ our card
  34. Wei Wu, Fei Wang, Arianna Yuan, Fei Wu, and Jiwei Li. 2020. CorefQA: Coreference resolution as query-based span prediction. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 6953–6963, Online. Association for Computational Linguistics. ↗ our card
  35. Juntao Yu, Bernd Bohnet, and Massimo Poesio. 2020. Neural mention detection. In Proceedings of the 12th Language Resources and Evaluation Conference, pages 1–10, Marseille, France. European Language Resources Association. ↗ our card
  36. Juntao Yu and Massimo Poesio. 2020. Multitask learning-based neural bridging reference resolution. In Proceedings of the 28th International Conference on Computational Linguistics, pages 3534–3546, Barcelona, Spain (Online). International Committee on Computational Linguistics. ↗ our card

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/hou-2021-end-end-neural-information-status/