What is it? Disambiguating the different readings of the pronoun 'it'

What it is about

English it can be anaphoric to a noun phrase («I have a bicycle. It is red»), event-referring to a verb phrase, clause or sentence («He lost his job. It came as a total surprise»), or pleonastic («It is raining»). Using 2,506 instances of it from the ParCor and DiscoMT corpora (52% anaphoric, 33% event, 15% pleonastic; event-reading this/that were substituted by it to get more data), the authors train a MaxEnt classifier on token context, the pronoun's head (with the complement of copular verbs instead of the verb), syntactic cues (a that-complement on the head or in the previous sentence, extraposed sentential subjects, nearest NP heads and adjective), and semantic cues (VerbNet selectional restrictions abstract/concrete; how often the head verb takes an event subject — a gerund or this — in Gigaword; a non-referential-it detector). Test accuracy is 68.7% (majority 53.9%): F1 .74 anaphoric, .74 pleonastic, .54 event. An RNN self-trained on silver labels does worse overall but better on events; the event class remains hardest.

Ideas

• I1. Classify every it three ways before the pronoun sieve runs, with rules from the paper's strongest cues, and let only the anaphoric ones reach nominal antecedents. This attacks Mova's worst pronoun (it: 41% precision on gold trees, 30% on SLM trees; the README's first error type). Pleonastic: weather and time verbs, it seems/appears/happens that, it is ADJ to/that… (extraposed clause in the tree), it takes; event: the head (or copular complement) is an event-taking predicate (happen, surprise, come as, be true, be a shame, mean, cause, make sense), or a that-clause / full sentence immediately precedes with no compatible noun. Output the class in coref explain; measure it/its precision and CoNLL F1 on GUM dev, gold and SLM trees.
• I2. Build the «takes an event subject» verb list from Mova's own parsed corpus, licence-clean, the way the paper used Gigaword: count, per verb lemma, how often its subject is a gerund, a clause, or this/that; keep verbs above a threshold. Store it as a data file next to prag/data/verbnet.tsv and use it in I1.
• I3. Send event-reading it/this/that to the previous event frame in the frame reader (prag/src/frames.rs): «He lost his job. It came as a surprise» → the patient of surprise is the losing frame, so «What surprised everyone?» can be answered. This is the frame-level half of the README's «event mentions and discourse deixis» item; measure frame bench coref and course what-questions about events.
• I4. Use VerbNet selectional restrictions already loaded in prag::vn: if the verb of the it-clause requires a concrete or animate subject, the event reading is out; if it requires an abstract one, nominal concrete antecedents are out.

For Mova

35. Reference and coreference §3.4 (abstract anaphora), §4 (non-referential uses), §9.3 (results: it/its). Code: coref/src/mention.rs and coref/src/sieve.rs (it classifier), prag/src/frames.rs (dummy it, event reference), prag/src/vn.rs. Measures: per-pronoun precision for it/its and CoNLL F1 on GUM dev; frame bench coref F1.

Topics: Coreference resolution · Semantic role labeling

Cite

The paper itself, not this card. BibTeX from the library's record of the work. DOI 10.18653/v1/d17-1137

@inproceedings{loaiciga2017what,
  title = {{What is it? Disambiguating the different readings of the pronoun 'it'}},
  author = {Loáiciga, Sharid and Guillou, Liane and Hardmeier, Christian},
  booktitle = {Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing},
  year = {2017},
  doi = {10.18653/v1/d17-1137},
  url = {https://doi.org/10.18653/v1/d17-1137}
}

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. Shane Bergsma and David Yarowsky. 2011. NADA: A robust system for non-referential pronoun detection. In Iris Hendrickx, Sobha Lalitha Devi, António Branco, and Ruslan Mitkov, editors, Anaphora Processing and Applications: 8th Discourse Anaphora and Anaphor Resolution Colloquium (DAARC), Lecture Notes in Artificial Intelligence, pages 12– 23. Springer, Faro, Portugal. find
2. Bernd Bohnet, Joakim Nivre, Igor Boguslavsky, Richárd Farkas, Filip Ginter, and Jan Hajič. 2013. Joint morphological and syntactic analysis for richly inflected languages. Transactions of the Association for Computational Linguistics, 1:415–428. find
3. Adriane Boyd, Whitney Gegg-Harrison, and Donna K. Byron. 2005. Identifying non-referential it: a machine learning approach incorporating linguistically motivated patterns. In Proceedings of the ACL Workshop on Feature Engineering for Machine Learning in Natural Language Processing, pages 40–47, Ann Arbor, Michigan. Association for Computational Linguistics. find
4. Donna K. Byron. 2002. Resolving pronominal reference to abstract entities. In Proceedings of the 40th Annual Meeting of the Association for Computational Linguistics, ACL 2002, pages 80–87, Philadelphia. Association for Computational Linguistics. find
5. Miriam Eckert and Michael Strube. 2000. Dialogue acts, synchronising units and anaphora resolution. Journal of Semantics, 17(1):51–89. find
6. Liane Guillou. 2015. Automatic post-editing for the DiscoMT pronoun translation task. In Proceedings of the Second Workshop on Discourse in Machine Translation, DiscoMT 2015, pages 65–71, Lisbon, Portugal. Association for Computational Linguistics. find
7. Liane Guillou. 2016. Incorporating Pronoun Function into Statistical Machine Translation. Ph.D. thesis, University of Edinburgh, Scotland, UK. find
8. Liane Guillou, Christian Hardmeier, Preslav Nakov, Sara Stymne, Jörg Tiedemann, Yannick Versley, Mauro Cettolo, Bonnie Webber, and Andrei Popescu-Belis. 2016. Findings of the 2016 WMT shared task on cross-lingual pronoun prediction. In Proceedings of the First Conference on Machine Translation, WMT16, pages 525–542, Berlin, Germany. Association for Computational Linguistics. find
9. Liane Guillou, Christian Hardmeier, Aaron Smith, Jörg Tiedemann, and Bonnie Webber. 2014. ParCor 1.0: A parallel pronoun-coreference corpus to support statistical MT. In Proceedings of the 9th International Conference on Language Resources and Evaluation, LREC 2014, pages 3191–3198, Reykjavik, Iceland. European Language Resources Association (ELRA). find
10. Christian Hardmeier, Jörg Tiedemann, Preslav Nakov, Sara Stymne, and Yannick Versely. 2016. DiscoMT 2015 Shared Task on Pronoun Translation. LINDAT/CLARIN digital library at Institute of Formal and Applied Linguistics, Charles University in Prague. find
11. Kenneth Heafield. 2011. KenLM: faster and smaller language model queries. In Proceedings of the Sixth Workshop on Statistical Machine Translation, WMT11, pages 187–197, Edinburgh, UK. Association for Computational Linguistics. find
12. Kailang Jiang, Giuseppe Carenini, and Raymond Ng. 2016. Training data enrichment for infrequent discourse relations. In Proceedings the 26th International Conference on Computational Linguistics: Technical Papers, COLING 2016, pages 2603– 2614, Osaka, Japan. find
13. The COLING 2016 Organizing Committee. find
14. Timothy Lee, Alex Lutz, and Jinho D. Choi. 2016. QA-It: classifying non-referential it for question answer pairs. In Proceedings of the ACL 2016 Student Research Workshop, pages 132–137, Berlin, Germany. Association for Computational Linguistics. find
15. Sharid Loáiciga, Liane Guillou, and Christian Hardmeier. 2016. It-disambiguation and source-aware language models for cross-lingual pronoun prediction. In Proceedings of the First Conference on Machine Translation, WMT16, pages 581–588, Berlin, Germany. Association for Computational Linguistics. find
16. Sharid Loáiciga, Thomas Meyer, and Andrei Popescu-Belis. 2014. English-French verb phrase alignment in Europarl for tense translation modeling. In Proceedings of the 9th International Conference on Language Resources and Evaluation, LREC 2014, pages 674–681, Reykjavik, Iceland. European Language Resources Association (ELRA). find
17. Juhani Luotolahti, Jenna Kanerva, and Filip Ginter. 2016. Cross-lingual pronoun prediction with deep recurrent neural networks. In Proceedings of the First Conference on Machine Translation, WMT16, pages 596–601, Berlin, Germany. Association for Computational Linguistics. find
18. Christopher Manning and Dan Klein. 2003. Max-Ent models, conditional estimation, and optimization without magic. Tutorial at HLT-NAACL and 41st ACL conferences. find
19. Christoph Müller. 2007. Resolving it, this, and that in unrestricted multi-party dialog. In Proceedings of the 45th Annual Meeting of the Association of Computational Linguistics, ACL 2007, pages 816–823, Prague, Czech Republic. Association for Computational Linguistics. find
20. Karin Kipper, Anna Korhonen, Neville Ryant, and Martha Palmer. 2008. A large scale classification of English verbs. Language Resources and Evaluation, 42(1):21–40. find
21. Courtney Napoles, Matthew Gormley, and Benjamin Van Durme. 2012. Annotated Gigaword. In Proceedings of the Joint Workshop on Automatic Knowledge Base Construction and Web-scale Knowledge Extraction, AKBC-WEKEX, pages 95–100, Montreal, Canada. Association for Computational Linguistics. find
22. Ronan Le Nagard and Philipp Koehn. 2010. Aiding pronoun translation with co-reference resolution. In Proceedings of the Joint 5th Workshop on Statistical Machine Translation, WMT10, pages 258–267, Uppsala, Sweden. Association for Computational Linguistics. find
23. Costanza Navarretta. 2004. Resolving individual and abstract anaphora in texts and dialogues. In Proceedings of the 20th International Conference on Computational Linguistics, COLING 2004, pages 233–239, Geneva, Switzerland. Association for Computational Linguistics. find
24. 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 Fifteenth Conference on Computational Natural Language Learning: Shared Task, CONLL 2011, pages 28–34, Portland, Oregon. Association for Computational Linguistics. find our card
25. Michal Novák, Anna Nedoluzhko, and Zdeněk Žabokrtský. 2013. Translation of “It” in a deep syntax framework. In Proceedings of the Workshop on Discourse in Machine Translation, DiscoMT 2015, pages 51–59, Sofia, Bulgaria. Association for Computational Linguistics. find our card
26. Princeton University. 2010. WordNet. find
27. Helmut Schmid. 1994. Probabilistic part-of-speech tagging using decision trees. In Proceedings of International Conference on New Methods in Language Processing, pages 44–49, Manchester, UK. find

This page in Markdown → https://md.mova.elamur.ai/free/en/docs/papers/loaiciga-2017-what-is-it-disambiguating-different/