Elamur Research⌕
☰
home / papers / 2018 / Towards Coreference for Literary Text: Analyzing Domain-Specific Phenomena

Towards Coreference for Literary Text: Analyzing Domain-Specific Phenomena

Summary

Automatic coreference resolvers are usually trained on news and dialogue data such as OntoNotes. Literary texts differ substantially from these genres, and existing literary coreference corpora were annotated without considering domain-specific issues. The authors carried out an iterative annotation study on thirteen literary works, mostly German, starting from existing guidelines and refining them as problems emerged. They identify phenomena that make coreference in fiction especially difficult:

  • the split between the narrator's world and the narrated world,
  • frequent shifts between generic and specific readings of noun phrases,
  • characters whose names, roles or identities change over the course of a story.

From these observations they propose guideline changes, such as annotating chains across entire books and allowing event antecedents. They also release a new keyboard-oriented annotation tool designed for long texts with many entities. The work matters for Digital Humanities tasks like character and relation analysis. It also raises the open question of how to represent, evaluate and learn from deliberately ambiguous references that may have no single correct annotation.

Key points

  • The study corpus contains 13 texts: Goethe's Werther, three Perutz novels, plays by Lessing and Schiller, six Grimm fairytales, and one 17th-century English captivity narrative by Mary Rowlandson.
  • Proposed guideline adaptations:
    • VPs, clauses and multi-sentence spans can serve as antecedents for events.
    • Group references (e.g., "they" covering two separately introduced people) are annotated.
    • NPs inside idioms are not annotated, and neither is bridging.
  • The new tool, CorefAnnotator, is open source under the Apache 2.0 license. It treats each coreference chain as an equivalence set, supports regex-based bulk annotation, and stores stand-off annotations via UIMA. It can import and export CoNLL 2012 and TEI/XML, and its interface is available in English and German.
  • The authors argue that intentional ambiguity in literature calls for corpora with multiple, conflicting annotations. Such corpora would challenge standard inter-annotator agreement measures and current training and evaluation practice.
  • As future work, they plan an openly available German literary corpus that will include works by Goethe and Perutz as well as several folktales.

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

11 papers of our library cite it (of the 140673 whose bibliographies we read); 5 of them have a page here, 6 not yet.

Cite

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

@inproceedings{roesiger2018coreference,
  title = {{Towards Coreference for Literary Text: Analyzing Domain-Specific Phenomena}},
  author = {Roesiger, Ina and Schulz, Sarah and Reiter, Nils},
  booktitle = {Proceedings of the Second Joint SIGHUM Workshop on Computational Linguistics for Cultural Heritage, Social Sciences, Humanities and Literature},
  year = {2018},
  url = {https://aclanthology.org/W18-4515/}
}

References

23 references, 8 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. R.T. Batista-Navarro and S. Ananiadou. 2011. Building a coreference-annotated corpus from the domain of biochemistry. find our card
  2. José Castaño, Jason Zhang, and James Pustejovsky. 2002. Anaphora resolution in biomedical literature. In Proceedings of the International Symposium on Reference Resolution for NLP. find
  3. Kevin Clark and Christopher D. Manning. 2016. Deep reinforcement learning for mention-ranking coreference models. In Empirical Methods on Natural Language Processing. htttp://uima.apache.org K. Bretonnel Cohen, Arrick Lanfranchi, William Corvey, William A. Baumgartner Jr., Christophe Roeder, Philip… find our card
  4. Stefanie Dipper, Anke Lüdeling, and Marc Reznicek. 2013. NoSta-D: A corpus of German non-standard varieties. Non-Standard Data Sources in Corpus-Based Research, (5):69–76. find
  5. George R. Doddington, Alexis Mitchell, Mark A. Przybocki, Lance A. Ramshaw, Stephanie Strassel, and Ralph M. Weischedel. 2004. The Automatic Content Extraction (ACE) Program - Tasks, Data, and Evaluation. In LREC. European Language Resources Association. find our card
  6. Caroline Gasperin and Ted Briscoe. 2008. Statistical anaphora resolution in biomedical texts. In Proceedings of the 22nd International Conference on Computational Linguistics-Volume 1, pages 257–264. Association for Computational Linguistics. find our card
  7. Caroline Gasperin, Nikiforos Karamanis, and Ruth Seal. 2007. Annotation of anaphoric relations in biomedical full-text articles using a domain-relevant scheme. In Proceedings of the 6th Discourse Anaphora and Anaphor Resolution Colloquium. find
  8. Evelyn Gius and Janina Jacke. 2017. The hermeneutic profit of annotation: On preventing and fostering disagreement in literary analysis. International Journal of Humanities and Arts Computing, 11(2):233–254, October. find
  9. Jan Hajič, Eduard Bejček, Alevtina Bémová, Eva Buráňová, Eva Hajičová, Jiřı́ Havelka, Petr Homola, Jiřı́ Kárnı́k, Václava Kettnerová, Natalia Klyueva, Veronika Kolářová, Lucie Kučová, Markéta Lopatková, Marie Mikulová, Jiřı́ Mı́rovský, Anna Nedoluzhko, Petr Pajas, Jarmila Pan… find
  10. Lynette Hirschman and Nancy Chinchor. 1998. Appendix F: MUC-7 Coreference Task Definition (version 3.0). In MUC. find our card
  11. G. Hirst. 1981. Anaphora in Natural Language Understanding: A Survey. Lecture Notes in Computer Science. Springer Berlin Heidelberg. find
  12. Markus Krug, Frank Puppe, Fotis Jannidis, Luisa Macharowsky, Isabella Reger, and Lukas Weimar. 2015. Rule-based coreference resolution in German historic novels. In Proceedings of the Fourth Workshop on Computational Linguistics for Literature, pages 98–104. find our card
  13. Uri Margolin. 1991. Reference, coreference, referring, and the dual structure of literary narrative. Poetics Today, 12(3):517–542. find
  14. M.A. Martı́, M. Taulé, M. Bertran, and L. Márquez. 2007. AnCora: Multilingual and Multilevel Annotated Corpora. find
  15. Karin Naumann and V Möller. 2006. Manual for the annotation of in-document referential relations. University of Tübingen. find
  16. 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. Association for Computational Linguistics. find our card
  17. Massimo Poesio and Ron Artstein. 2008. Anaphoric Annotation in the ARRAU Corpus. In International Conference on Language Resources and Evaluation (LREC), Marrakech, Morocco, May. find our card
  18. Massimo Poesio, Jon Chamberlain, Udo Kruschwitz, Livio Robaldo, and Luca Ducceschi. 2013. Phrase detectives: Utilizing collective intelligence for internet-scale language resource creation. ACM Transactions on Interactive Intelligent Systems (TiiS), 3(1):3. find
  19. T. Robbins. 1976. Even cowgirls get the blues. Bantam Books: Novel. Bantam Books. find
  20. Ina Rösiger and Simone Teufel. 2014. Resolving coreferent and associative noun phrases in scientific text. In Proceedings of EACL. find
  21. Olga Uryupina, Ron Artstein, Antonella Bristot, Ferederica Cavicchio, Francesca Delogu, Kepa Rodriguez, and Massimo Poesio. to appear. Annotating a broad range of anaphoric phenomena, in a variety of genres: the ARRAU Corpus. Journal of Natural Language Engineering. find
  22. Andreas van Cranenburgh and Rens Bod. 2017. A data-oriented model of literary language. In Proceedings of EACL. find
  23. Ralph Weischedel, Eduard Hovy, Mitchell Marcus, Martha Palmer, Robert Belvin, Sameer Pradhan, Lance Ramshaw, and Nianwen Xue. 2011. Ontonotes: A large training corpus for enhanced processing. pages 54–63. find

↑ back to top