Towards Named-Entity and Coreference Annotation of the Hebrew Bible

What it is about

The authors extend the Ancient Hebrew UD treebank (UD_Ancient_Hebrew-PTNK) with named-entity and coreference annotation, using the CorefUD format, so that each mention points to an entry in a separate entity list rather than to an earlier phrase. The coverage so far is the first 40 chapters of Genesis: about 1,500 entities and over 10,000 mentions. Mention rules are taken from the English GUM guidelines: every noun phrase, proper noun and pronoun, including nested ones, counts, with interrogatives and a few non-referential constructions excluded. There are ten entity types, given one-letter codes (Person, Location, Abstract, Inanimate, Creature, Nation, Time, Substance, Vegetation, Event). A rule-based coreferencer was too inaccurate to keep, so a person annotates in a simple terminal tool. Genesis 6 was annotated twice: span agreement was 88.6% (93.1% after correction), coreference agreement 87.8%, and entity-type agreement 82.3%. The averaged CoNLL F1 across the metrics is 72.55. Persons make up about 70% of the mentions, and eleven entities account for 37% of them.

Ideas

• I1. A rule-based mention detector for macula prag: every noun phrase and pronoun in the trees is a candidate, with the GUM exclusions as rules, and an annotator confirms or rejects each candidate. This gives the rule-based pragmatics a coreference layer without a learned coreferencer.
• I2. Keep a name apart from its referent: the paper gives "Isaac" as a name a separate entity from Isaac himself. In world, a name should link to its referent as a second entity, not merge with it.
• I3. Measure agreement the way the paper does: double-annotate one chapter and report the CoNLL-style averages with the corefUD scorer. Use this as the gold check for any coreference layer added to treebanks/hbo/ptnk.

For Mova

I1 is a rule for macula prag. I2 is a design point for the entities in prag and world. I3 is a measure for the treebank layer in treebanks/hbo/ptnk. The ten entity types are a candidate tag set for the lexicon notes in The lexicon.

Topics: Named entity recognition · Coreference resolution · Treebanks

Cite

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

@inproceedings{swanson2024named,
  title = {{Towards Named-Entity and Coreference Annotation of the Hebrew Bible}},
  author = {Swanson, Daniel G. and Bussert, Bryce D. and Tyers, Francis},
  booktitle = {Proceedings of the Third Workshop on Language Technologies for Historical and Ancient Languages (LT4HALA) @ LREC-COLING-2024},
  year = {2024},
  url = {https://aclanthology.org/2024.lt4hala-1.5/}
}

References

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

1. Jacob Cohen. 1960. A coefficient of agreement for nominal scales. Educational and Psychological Measurement, 20(1):37–46. ↗
2. Anna Nedoluzhko, Michal Novák, Martin Popel, Zdeněk Žabokrtský, Amir Zeldes, and Daniel Zeman. 2022. CorefUD 1.0: Coreference meets Universal Dependencies. In Proceedings of the Thirteenth Language Resources and Evaluation Conference, pages 4859–4872, Marseille, France. European Language Resources Association. ↗ our card
3. Joakim Nivre, Marie-Catherine de Marneffe, Filip Ginter, Jan Hajic, Christopher D Manning, Sampo Pyysalo, Sebastian Schuster, Francis Tyers, and Daniel Zeman. 2020. Universal dependencies v2: An evergrowing multilingual treebank collection. In Proceedings of the 12th Language Resources and Evaluation Conference, pages 4034–4043. find our card
4. W.T. van Peursen, C. Sikkel, and D. Roorda. 2015. Hebrew text database ETCBC4b. find
5. Daniel Swanson and Francis Tyers. 2022. A Universal Dependencies treebank of Ancient Hebrew. In Proceedings of the Thirteenth Language Resources and Evaluation Conference, pages 2353–2361, Marseille, France. European Language Resources Association. find our card
6. Juntao Yu, Sopan Khosla, Nafise Sadat Moosavi, Silviu Paun, Sameer Pradhan, and Massimo Poesio. 2022. The universal anaphora scorer. In Proceedings of the Thirteenth Language Resources and Evaluation Conference, pages 4873–4883, Marseille, France. European Language Resources Association. find
7. Amir Zeldes. 2017. The GUM corpus: Creating multilayer resources in the classroom. Language Resources and Evaluation, 51(3):581–612. ↗
8. Amir Zeldes and Shuo Zhang. 2016. When annotation schemes change rules help: A configurable approach to coreference resolution beyond OntoNotes. In Proceedings of the Workshop on Coreference Resolution Beyond OntoNotes (CORBON 2016), pages 92–101, San Diego, California. Association for Computational Linguistics. find

This page in Markdown → https://md.mova.elamur.ai/free/en/docs/papers/coref-2024-swanson-named-entity-coreference-annotation-hebrew/