Elamur Research⌕
☰
home / papers / 2022 / Building a Manually Annotated Hungarian Coreference Corpus: Workflow and Tools

Building a Manually Annotated Hungarian Coreference Corpus: Workflow and Tools

Summary

This paper tackles the shortage of manually checked coreference data for Hungarian, a pro-drop language. It introduces KorKor, a small corpus built from Hungarian Wikipedia and GlobalVoices texts taken from OPUS. The texts were processed with the emtsv pipeline, and annotators reviewed the output in three rounds across the workflow, so every layer was hand-corrected: lemmas, POS tags, morphology (two tagsets), dependencies, and anaphora/coreference. The scheme differs from the earlier SzegedKoref. It uses only two nominal relation tags but several fine-grained pronominal ones, and it adds zero elements, including covert subjects of infinitives and elided verbs. The corpus comes in two versions. The xtsv version keeps zero nodes and branching reference networks, which suits linguistic study. The CoNLL-U Plus version drops zero tokens and links mentions in simple chains, which makes it better for training and testing resolvers. The main contribution is the openly licensed corpus together with its reusable workflow, guidelines and tools. The author also raises unresolved problems such as split antecedents and referents whose state changes.

Key points

  • Size: KorKor has 94 documents (62 Wikipedia, 32 GlobalVoices). The CoNLL-U Plus version has 24,499 tokens and the xtsv version has 26,581. Another 8,600 tokens have corrected morphology and dependencies but no coreference layer.
  • License and access: The corpus is released under CC-BY 4.0 on GitHub (vadno/korkor_pilot), together with the workflow instructions, guidelines and tools.
  • Tagset: Nominal relations use only two tags, identity and part-whole (holonym). Pronominal anaphora is split by pronoun type (personal, demonstrative, reciprocal, reflexive, possessive). Extra tags cover generic subjects, speaker and addressee.
  • Zero elements: Dropped subjects, objects and possessors are annotated, as are zero substantive verbs and ellipted verbs. In the CoNLL-U Plus version, dropped pronouns are stored in a separate column and excluded from coreference chains.
  • Comparison: SzegedKoref, the largest Hungarian coreference corpus, has 55,763 tokens and 2,456 chains. It was the design inspiration but uses a different relation scheme.

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 · Universal Dependencies · Part-of-speech tagging

Cited by

5 papers of our library cite it (of the 140673 whose bibliographies we read); 4 of them have a page here, 1 not yet.

Cite

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

@inproceedings{vadasz2022building,
  title = {{Building a Manually Annotated Hungarian Coreference Corpus: Workflow and Tools}},
  author = {Vadász, Noémi},
  booktitle = {Proceedings of the Fifth Workshop on Computational Models of Reference, Anaphora and Coreference},
  year = {2022},
  doi = {10.48448/4x6j-qa08},
  url = {https://doi.org/10.48448/4x6j-qa08}
}

References

16 references, 3 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. J. Baptista, Simone Pereira, and Nuno J. Mamede. 2016. Zac : Zero anaphora corpus a corpus for zero anaphora resolution in portuguese. In Proceedings of Workshop on Corpora and Tools for Processing Corpora, PROPOR 2016. find
  2. Attila Novák. 2003. Milyen a jó Humor? (What good humor is like?). In I. Magyar Számítógépes Nyelvészeti Konferencia, pages 138–144, Szeged. SZTE. find
  3. Dóra Csendes, János Csirik, Tibor Gyimóthy, and András Kocsor. 2005. The Szeged Treebank. In Proceedings of the 8th International Conference, TSD 2005, pages 123–131, Karlovy Vary, Czech Republic. Springer. find
  4. Attila Novák. 2014. A new form of Humor – Mapping constraint-based computational morphologies to a finite-state representation. In Proceedings of the Ninth International Conference on Language Resources and Evaluation (LREC’14), Reykjavik, Iceland. European Language Resources Association (ELRA). find
  5. Richard Eckart de Castilho, Éva Mújdricza-Maydt, Seid Muhie Yimam, Silvana Hartmann, Iryna Gurevych, Anette Frank, and Chris Biemann. 2016. A web-based tool for the integrated annotation of semantic and syntactic structures. In Proceedings of the Workshop on Language Technology Resources Attila Nová… find
  6. Maciej Ogrodniczuk, Katarzyna Głowińska, Mateusz Kopeć, Agata Savary, and Magdalena Zawisławska. 2016. Polish coreference corpus. In Human Language Technology. Challenges for Computer Science and Linguistics, pages 215–226, Cham. Springer International Publishing. find
  7. György Orosz and Attila Novák. 2012. PurePos 2.0 – an open source morphological disambiguator. In Proceedings of the 9th International Workshop on Natural Language Processing and Cognitive Science, pages 53–63, Wroclaw. find
  8. György Orosz and Attila Novák. 2013. PurePos 2.0: a hybrid tool for morphological disambiguation. In Proceedings of the International Conference Recent Advances in Natural Language Processing RANLP 2013, pages 539–545, Hissar, Bulgaria. INCOMA Ltd. Shoumen, BULGARIA. find
  9. Sameer Pradhan, Alessandro Moschitti, Nianwen Xue, Olga Uryupina, and Yuchen Zhang. 2012. CoNLL-2012 shared task: Modeling multilingual unrestricted coreference in OntoNotes. In Joint Conference on EMNLP and CoNLL - Shared Task, pages 1–40, Jeju Island, Korea. Association for Computational Linguisti… find our card
  10. Marta Recasens and Antonia Martí. 2010. Ancora-co: Coreferentially annotated corpora for spanish and catalan. Language Resources and Evaluation, 44:315–345. find
  11. Jörg Tiedemann. 2012. Parallel data, tools and interfaces in opus. In Proceedings of the Eight International Conference on Language Resources and Evaluation (LREC’12), Istanbul, Turkey. European Language Resources Association (ELRA). find
  12. Noémi Vadász and Eszter Simon. 2019. Konverterek magyar morfológiai címkekészletek között (Converters between Hungarian morphological tagsets). In XV. find
  13. Magyar Számítógépes Nyelvészeti Konferencia (MSZNY 2019), pages 99–112, Szeged. Szegedi Tu-dományegyetem Informatikai Tanszékcsoport. find
  14. Kees van Deemter and Rodger Kibble. 1999. What is coreference, and what should coreference annotation be? In Coreference and Its Applications, pages 90– 96. find our card
  15. Veronika Vincze, Klára Hegedűs, Alex Sliz-Nagy, and Richárd Farkas. 2018. SzegedKoref: A Hungarian coreference corpus. In Proceedings of the 11th Language Resources and Evaluation Conference, Miyazaki, Japan. European Language Resource Association. find our card
  16. Ralph Weischedel, Martha Palmer, Mitchell Marcus, Eduard Hovy, Sameer Pradhan, Lance Ramshaw, Nianwen Xue, Ann Taylor, Jeff Kaufman, Michelle Franchini, Mohammed El-Bachouti, Robert Belvin, and Ann Houston. 2013. OntoNotes Release 5.0. find

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/papers/vadasz-2022-building-manually-annotated-hungarian-coreference/