Elamur Research⌕
☰
home / papers / 2004 / The NomBank Project: An Interim Report

The NomBank Project: An Interim Report

Summary

NomBank tackles a gap in corpus resources. Verbs in the Penn Treebank II already had predicate-argument annotation through PropBank, but common nouns did not. The project labels argument structure for about 5000 noun predicates, reusing PropBank's numbered roles (ARG0–ARG4), its adjunct types (ARGM-*) and its tooling. It adds SUPPORT items: verbs such as "made" in "made a decision" that connect arguments outside the NP to the noun.

To prepare the work, the team merged COMLEX Syntax, NOMLEX and Levin's verb classes into a new dictionary, NOMLEX-PLUS. They used it to sort nouns into classes and to generate draft frames automatically. Some drafts were converted from PropBank verb entries, others came from class templates, and annotators then corrected them.

Quality control relies on automatic checks that send only suspicious instances back for a second look. This interim report describes the specifications, the workflow and planned future work. Because nouns and verbs share one role scheme, a system can treat phrasings like "IBM appointed John" and "IBM's appointment of John" as the same event. Noun argument structure had previously received little attention.

Key points

  • Target: argument structure for roughly 5000 common nouns in Penn Treebank II. Labels are REL, SUPPORT, ARG0–ARG4 and ARGM types, shared with PropBank.
  • NOMLEX-PLUS holds the original 1000 NOMLEX entries plus 6000 new ones. Its classes include verbal nominalizations, adjectival nominalizations and 16 further classes, such as relational and partitive nouns.
  • Staffing: one supervisor and one annotator for the first nine months while the specifications were settled, later growing to four annotators.
  • Error detection uses several automatic checks:
    • argument nominalizations checked against NOMLEX-PLUS (e.g., acquirer = ARG0)
    • required and illegal labels
    • unlikely role configurations, such as a duplicated role
    • wrong-POS detection via GLARF tools
  • About 5–10% of instances are flagged for a second pass. Planned next is a rule-based PROPBANKER for automatic annotation, in contrast to statistical systems like Gildea and Jurafsky (2002).

Interesting ideas

  • I1. "À la carte" layered annotation: each linguistic layer is defined by its own annotated data over the same corpus. Tools for mapping between layers can then be built and swapped independently, which would also suit a growing multi-layer Ukrainian corpus.
  • I2. Bootstrapping frames from related resources: draft noun entries were derived automatically from verb frames, and from "cousin" verbs that are semantically similar but morphologically unrelated (e.g., advantage modeled on exploit). Humans only edit the drafts, which saves lexicographic effort when extending a resource to a new word class.
  • I3. Class-level template frames: for nouns with no verbal source, one frame is written per noun class and inherited by every member, with per-word fixes. This cheaply covers a long tail of vocabulary.
  • I4. Linguistically motivated sanity checks: constraints like "the same role should not appear twice", or rare positional patterns, flag likely errors. Re-checking only that small flagged subset keeps quality control affordable.
  • I5. Imperfect automation still has value: even an automatic annotator too weak to serve as a preprocessor can expose human errors, because disagreements with the manual labels point to instances worth reviewing.

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 · Natural language processing

Topics: Semantic role labeling

Cited by

136 papers of our library cite it (of the 140673 whose bibliographies we read); 20 of them have a page here, 116 not yet.

Cite

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

@inproceedings{meyers2004nombank,
  title = {{The NomBank Project: An Interim Report}},
  author = {Meyers, Adam and Reeves, Ruth and Macleod, Catherine and Szekely, Rachel and Zielinska, Veronika and Young, Brian and Grishman, Ralph},
  booktitle = {Proceedings of the Workshop Frontiers in Corpus Annotation at HLT-NAACL 2004},
  year = {2004},
  url = {https://aclanthology.org/W04-2705/}
}

References

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

  1. T. Fontenelle. 1997. Turning a bilingual dictionary into a lexical-semantic database. Lexicographica Series Maior 79. Max Niemeyer Verlag, Tübingen. ↗
  2. D. Gildea and D. Jurafsky. 2002. Automatic Labeling of Semantic Roles. Computational Linguistics, 28:245– 288. ↗ our card
  3. M. Gross. 1981. Les bases empiriques de la notion de prédicat sémantique. In A. Guillet and C. Lecl‘ere, editors, Formes Syntaxiques et Prédicat Sémantiques, volume 63 of Langages, pages 7–52. Larousse, Paris. ↗
  4. M. Gross. 1982. Simple Sentences: Discussion of Fred W. Householder’s Paper “Analysis, Synthesis and Improvisation”. In Text Processing. Text Analysis and Generation. Text Typology and Attribution. Proceedings of Nobel Symposium 51. find
  5. N. Habash and B. Dorr. 2003. CatVar: A Database of Categorial Variations for English. In Proceedings of the MT Summit, pages 471–474, New Orleans. find our card
  6. P. Kingsbury and M. Palmer. 2002. From treebank to propbank. In Proceedings of the 3rd International Conference on Language Resources and Evaluation (LREC-2002), Las Palmas, Spain. find our card
  7. P. Kingsbury, M. Palmer, and Mitch Marcus. 2002. Adding semantic annotation to the penn treebank. In Proceedings of the Human Language Technology Conference, San Diego, California. find
  8. B. Levin. 1993. English Verb Classes and Alternations: A Preliminary Investigation. The University of Chicago Press, Chicago. find
  9. C. Macleod, R. Grishman, and A. Meyers. 1998a. COMLEX Syntax. Computers and the Humanities, 31(6):459–481. find
  10. C. Macleod, R. Grishman, A. Meyers, L. Barrett, and R. Reeves. 1998b. Nomlex: A lexicon of nominalizations. In Proceedings of Euralex98. find
  11. I. A. Mel’ĉuk. 1988. Dependency Syntax: Theory and Practice. State University Press of New York, Albany. find
  12. I. A. Mel’ĉuk. 1996. Lexical Functions: A Tool for the Description of Lexical Relations in a Lexicon. In Lexical Functions in Lexicography and Natural Language Processing. John Benjamins Publishing Company, Amsterdam. find
  13. A. Meyers, C. Macleod, R. Yangarber, R. Grishman, Leslie Barrett, and Ruth Reeves. 1998. Using NOM-LEX to Produce Nominalization Patterns for Information Extraction. In Coling-ACL98 workshop Proceedings: the Computational Treatment of Nominals. find
  14. A. Meyers, R. Grishman, M. Kosaka, and S. Zhao. 2001a. Covering Treebanks with GLARF. In ACL/EACL Workshop on Sharing Tools and Resources for Research and Education. find
  15. A. Meyers, M. Kosaka, S. Sekine, R. Grishman, and S. Zhao. 2001b. Parsing and GLARFing. In Proceedings of RANLP-2001, Tzigov Chark, Bulgaria. find
  16. A. Meyers, R. Reeves, Catherine Macleod, Rachel Szeke-ley, Veronkia Zielinska, Brian Young, and R. Grishman. 2004. The Cross-Breeding of Dictionaries. In Proceedings of LREC-2004, Lisbon, Portugal. To appear. find
  17. MUC-6. 1995. Proceedings of the Sixth Message Understanding Conference. Morgan Kaufman. (MUC-6). find
  18. D. M. Perlmutter and P. M. Postal. 1984. The 1-Advancement Exclusiveness Law. In D. M. Perlmutter and C. G. Rosen, editors, Studies in Relational Grammar 2. The University of Chicago Press, Chicago. find
  19. M. Poesio and R. Vieira. 1998. A Corpus-based Investigation of Definite Description Use. Computational Linguistics, 24(2):183–216. ↗
  20. C. G. Rosen. 1984. The Interface between Semantic Roles and Initial Grammatical Relations. In D.. M. Perlmutter and C. G. Rosen, editors, Studies in Relational Grammar 2. The University of Chicago Press, Chicago. find
  21. University of Pennsylvania. 2002. Annotation guidelines for PropBank. http://www.cis.upenn.edu/˜ace/propbank-guidelines-feb02.pdf. ↗

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/meyers-2004-nombank-project-interim-report/