Mova0.5 pre-release⌕
☰
home / docs / papers / English / Domain-Aware Dependency Parsing for Questions

Domain-Aware Dependency Parsing for Questions

What it is about

Pre-trained dependency parsers are trained on corpora where questions are rare (0.5% of Wall Street Journal sentences), so they fail on domain questions such as "Show me Neil's insider transactions since 2011". The authors generate labelled domain questions from a few seed questions that carry UD parses. Each seed becomes a template in which the subject and modifier entities are placeholders, and the templates are refilled from a small domain schema and knowledge base. A one-time schema annotation (connective prepositions, heading properties, class-dependency, possible wh-words) removes the unnatural outputs. Each domain has 250 hand-parsed questions, evaluated with 5-fold cross-validation. Training on the generated questions raises LAS by 4.9% to 9% over UD plus the general QuestionBank, and 10 seed questions already give 2.03% (Finance) and 2.82% (Weather) LAS gains with Malt. The schema annotations add 0.4% to 0.94% LAS.

Ideas

For Mova

The question side belongs to 26. Questions; the whose-expansion of a modifier fits 25. Relative clauses; the temporal modifier fits 07. Punctuation, capital letters, abbreviations and numbers only for dates. Code: en/src/parse.rs (the attachment errors), world/src/knowledge.rs (schema annotations as school course data), world/src/qa.rs (question forms). Measure: UAS and LAS on a held-out set of school-domain questions, split by template, with the UD EWT dev score as a check that nothing regressed.

Topics: Dependency parsing

Cite

The paper itself, not this card. BibTeX from the library's record of the work. DOI 10.18653/v1/2021.findings-acl.400

@inproceedings{garimella2021domain,
  title = {{Domain-Aware Dependency Parsing for Questions}},
  author = {Garimella, Aparna and Chiticariu, Laura and Li, Yunyao},
  booktitle = {Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021},
  year = {2021},
  doi = {10.18653/v1/2021.findings-acl.400},
  url = {https://doi.org/10.18653/v1/2021.findings-acl.400}
}

References

12 references, 2 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Daniel Andor, Chris Alberti, David Weiss, Aliaksei Severyn, Alessandro Presta, Kuzman Ganchev, Slav Petrov, and Michael Collins. 2016. Globally normalized transition-based neural networks. In Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), ACL 2016, pages 2442–2452, Berlin, Germany. find
  2. Danica Damljanovic, Milan Agatonovic, and Hamish Cunningham. 2010. Natural language interfaces to ontologies: Combining syntactic analysis and ontology-based lookup through the user interaction. In Proceedings of the 7th International Conference on The Semantic Web: Research and Applications - Volume Part I, ESWC’10, pages 106–120, Heraklion, Crete, Greece. find
  3. Jason M Eisner. 1996. Three new probabilistic models for dependency parsing: An exploration. In Proceedings of the 16th conference on Computational linguistics-Volume 1, COLING 1996, pages 340– 345, Copenhagen, Denmark. find our card
  4. Thierry Hamon, Natalia Grabar, and Fleur Mougin. 2017. Querying biomedical linked data with natural language questions. Semantic Web, 8(4):581–599. find
  5. John Judge, Aoife Cahill, and Josef Van Genabith. 2006. QuestionBank: Creating a corpus of parse-annotated questions. In Proceedings of the 21st International Conference on Computational Linguistics and the 44th annual meeting of the Association for Computational Linguistics, ACL 2006, pages 497–504, Sydney. find
  6. Emilie Pasche Dina Vishnyakova Pascale Gaudet Amos Bairoch Julien Gobeill, Arnaud Gaudinat and Patrick Ruch. 2015. Deep question answering for protein annotation. Database: The Journal of Biological Databases and Curation, bav081. find
  7. Joakim Nivre, Johan Hall, and Jens Nilsson. 2004. Memory-based dependency parsing. In Proceedings of the Eighth Conference on Computational Natural Language Learning (CoNLL-2004) at HLT-NAACL 2004, New York City, USA. find
  8. Joakim Nivre, Johan Hall, and Jens and Nilsson. 2016a. Malt Parser, v1.9, released June 4 2016. http:// www.maltparser.org/download.html . ↗
  9. Joakim Nivre, Marie-Catherine de Marneffe, Filip Ginter, Yoav Goldberg, Jan Hajic, Christopher D. Manning, Ryan McDonald, Slav Petrov, Sampo Pyysalo, Natalia Silveira, Reut Tsarfaty, and Daniel Zeman. 2016b. Universal dependencies v1: A multilingual treebank collection. In Proceedings of the Tenth International Conference on Language Resources and Evaluation, LREC 2016, pages 1659–1666, Portorož, Slovenia. find our card
  10. Slav Petrov, Pi-Chuan Chang, Michael Ringgaard, and Hiyan Alshawi. 2010. Uptraining for accurate deterministic question parsing. In Proceedings of the 2010 Conference on Empirical Methods in Natural Language Processing, EMNLP 2010, pages 705– 713, Cambridge, MA, USA. find
  11. Natalia Silveira, Timothy Dozat, Marie-Catherine de Marneffe, Samuel Bowman, Miriam Connor, John Bauer, and Christopher D. Manning. 2014. A gold standard dependency corpus for English. In Proceedings of the Ninth International Conference on Language Resources and Evaluation, LREC 2014, pages 2897–2904, Reykjavik, Iceland. find
  12. Yushi Wang, Jonathan Berant, and Percy Liang. 2015. Building a semantic parser overnight. 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), ACL IJCNLP 2015, pages 1332–1342, Beijing, China. find

This page in Markdown → https://md.mova.elamur.ai/free/en/docs/papers/garimella-2021-domain-aware-dependency-parsing-questions/