Adding Noun Phrase Structure to the Penn Treebank

What it is about

The Penn Treebank leaves base noun phrases flat, so "Air Force contract" has no internal structure, and any parser trained on it cannot learn that "Air Force" is a unit. The authors add gold bracketing inside every ambiguous noun phrase, using NML nodes for nominal modifiers and JJP nodes for adjective phrases, with guidelines, named-entity support and a check of inter-annotator agreement. Of the noun phrases, 88.65 percent are unchanged from the original, and the new brackets agree with the sub-NP structure of DepBank at an F-score of 85.72. Two uses are tested: NP bracketing, where the dependency and adjacency models behave differently on the new corpus than on Lauer's set, and Collins parsing, where adding the new brackets lowers the overall F-score only from 88.92 to 88.46.

Ideas

• I1. Nominal-modifier split in en/src/convert.rs: for PTB-derived banks, emit compound for a noun-internal modifier and amod for an adjectival one, following the NML and JJP split; measure compound and amod F1 on EWT dev, since EWT comes from the PTB.
• I2. A sample check: take 100 multiword noun phrases from EWT dev, label their internal structure by the NML and JJP scheme, and measure how often en picks the same head; this is a measurement, not training data.
• I3. Own NP-bracketing count model: with counts from Mova's own silver corpus, compare the adjacency and dependency scores of the paper's Table 6 on noun compounds, before any statistical NP-bracketing component is added to prag (Lauer's data, test only if its licence allows).

For Mova

The converter is en/src/convert.rs and the noun-phrase side of the parser is en/src/parse.rs. The chapter is 22. Phrases: noun, adjective, adverb and prepositional phrases, and 19. The lexicon: vocabulary, multiword expressions and register for multiword nouns. Measure first the compound and amod F1 on EWT dev, which shows whether the split changes anything.

Topics: Treebanks

Cite

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

@inproceedings{vadas2007adding,
  title = {{Adding Noun Phrase Structure to the Penn Treebank}},
  author = {Vadas, David and Curran, James},
  booktitle = {Proceedings of the 45th Annual Meeting of the Association of Computational Linguistics},
  year = {2007},
  url = {https://aclanthology.org/P07-1031/}
}

References

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

1. Ann Bies, Mark Ferguson, Karen Katz, and Robert MacIntyre. 1995. Bracketing guidelines for Treebank II style Penn Treebank project. Technical report, University of Pennsylvania. find
2. Dan Bikel. 2004. On the Parameter Space of Generative Lexicalized Statistical Parsing Models. Ph.D. thesis, University of Pennsylvania. find
3. Thorsten Brants and Alex Franz. 2006. Web 1T 5-gram version 1. Linguistic Data Consortium. find
4. Ted Briscoe and John Carroll. 2006. Evaluating the accuracy of an unlexicalized statistical parser on the PARC DepBank. In Proceedings of the Poster Session of COLING/ACL-06. Sydney, Australia. find
5. Michael Collins. 1999. Head-Driven Statistical Models for Natural Language Parsing. Ph.D. thesis, University of Pennsylvania. find our card
6. Usama M. Fayyad and Keki B. Irani. 1993. Multi-interval discretization of continuous-valued attributes for classification learning. In Proceedings of the 13th International Joint Conference on Artifical Intelligence (IJCAI–93), pages 1022– 1029. Chambery, France. find
7. Roxana Girju, Dan Moldovan, Marta Tatu, and Daniel Antohe. 2005. On the semantics of noun compounds. Journal of Computer Speech and Language - Special Issue on Multiword Expressions, 19(4):313–330. find
8. Julia Hockenmaier. 2003. Data and Models for Statistical Parsing with Combinatory Categorial Grammar. Ph.D. thesis, University of Edinburgh. find
9. Tracy Holloway King, Richard Crouch, Stefan Riezler, Mary Dalrymple, and Ronald M. Kaplan. 2003. The PARC700 dependency bank. In Proceedings of the 4th International Workshop on Linguistically Interpreted Corpora (LINC-03). Budapest, Hungary. find
10. Seth Kulick, Ann Bies, Mark Libeman, Mark Mandel, Ryan McDonald, Martha Palmer, Andrew Schein, and Lyle Ungar. 2004. Integrated annotation for biomedical information extraction. In Proceedings of the Human Language Technology Conference of the North American Chapter of the Association for Computational Linguistics. Boston. find
11. Mirella Lapata and Frank Keller. 2004. The web as a baseline: Evaluating the performance of unsupervised web-based models for a range of NLP tasks. In Proceedings of the Human Language Technology Conference of the North American Chapter of the Association for Computational Linguistics, pages 121–128. Boston. find
12. Mark Lauer. 1995. Corpus statistics meet the compound noun: Some empirical results. In Proceedings of the 33rd Annual Meeting of the Association for Computational Linguistics. Cambridge, MA. find
13. Mitchell Marcus. 1980. A Theory of Syntactic Recognition for Natural Language. MIT Press, Cambridge, MA. find
14. Mitchell Marcus, Beatrice Santorini, and Mary Marcinkiewicz. 1993. Building a large annotated corpus of English: The Penn Treebank. Computational Linguistics, 19(2):313–330. find our card
15. Preslav Nakov and Marti Hearst. 2005. Search engine statistics beyond the n-gram: Application to noun compound bracketing. In Proceedings of CoNLL-2005, Ninth Conference on Computational Natural Language Learning. Ann Arbor, MI. find
16. Lance A. Ramshaw and Mitchell P. Marcus. 1995. Text chunking using transformation-based learning. In Proceedings of the Third ACL Workshop on Very Large Corpora. Cambridge MA, USA. find
17. Mark Steedman. 2000. The Syntactic Process. MIT Press, Cambridge, MA. find
18. Ralph Weischedel and Ada Brunstein. 2005. BBN pronoun coreference and entity type corpus. Technical report, Linguistic Data Consortium. find

This page in Markdown → https://md.mova.elamur.ai/free/en/docs/papers/bank-2007-vadas-adding-noun-phrase-structure-penn/