Elamur Research⌕
☰
home / papers / 2012 / The Dependency-Parsed FrameNet Corpus

The Dependency-Parsed FrameNet Corpus

Summary

FrameNet 1.5 provides manual frame-semantic annotation for more than 170,000 English sentences, mostly drawn from the British National Corpus. It has no standard syntactic analysis. Its frame elements are marked as surface text spans without head words. Semantic role labeling (SRL) researchers therefore parse the sentences themselves and align the parses with the annotations. This alignment step is rarely documented, which makes results hard to reproduce and compare.

The authors parse every sentence with the Stanford Parser, producing Stanford Dependencies. They then map each frame-evoking element and each frame element onto a single dependency node. A frame element's head is chosen as the node whose subtree best fits the annotated tokens, scored with a weighted F-measure of precision and recall. The same score also selects the most compatible parse from the parser's n-best list.

Tested against manually annotated heads, the method recovers them with close to 91% accuracy. The output is released as a public resource intended to standardize FrameNet-based SRL work.

Key points

  • Tuning showed that the best setting relies on recall alone (β = ∞). Considering the 30 best parses gave the largest gain, and adding more parses brought no further improvement.
  • The corpus comes in two variants, one using basicDependencies and one using CCPropagatedDependencies. The collapsed variant omits frames evoked by prepositions.
  • Data use a CoNLL-2008-style column format with TreeTagger lemmas and POS tags, the original FrameNet POS tags, and one pair of FEE/FE columns per frame.
  • Each sentence records the rank of the selected parse, its mapping score, and a shuffled random ID (RID) for reproducible subsets.
  • The release includes the Das and Smith (2011) train/test split as stand-off annotations and a script that realigns the data with the original FrameNet text for evaluation.

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: Semantic role labeling

Cite

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

@inproceedings{bauer2012dependency,
  title = {{The Dependency-Parsed FrameNet Corpus}},
  author = {Bauer, Daniel and Fürstenau, Hagen and Rambow, Owen},
  booktitle = {Proceedings of the Eighth International Conference on Language Resources and Evaluation (LREC'12)},
  year = {2012},
  doi = {10.63317/32zub9vntbgb},
  url = {https://doi.org/10.63317/32zub9vntbgb}
}

References

16 references, 6 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. C. Baker, M. Ellsworth, and K. Erk. 2007. Semeval’07 task 19: frame semantic structure extraction. In Proceedings HEAD 2 find
  2. DEPREL nsubj FRAME FE Agent Body movement 4 2 2 5 poss dobj prep pobj Body part Internal cause of the 4th International Workshop on Semantic Evaluations, pages 99–104. Association for Computational Linguistics. find
  3. Dipanjan Das and Noah A. Smith. 2011. Semi-supervised frame-semantic parsing for unknown predicates. In Proceedings of the 49th Meeting of the Association for Computational Linguistics, pages 1435–1444. find our card
  4. Marie-Catherine de Marneffe, Bill MacCartney, and Christopher D. Manning. 2006. Generating typed dependency parses from phrase structure parses. In Proceedings of LREC 2006. find our card
  5. Charles J. Fillmore, Christopher R. Johnson, and Miriam R. L. Petruck. 2003. Background to FrameNet. International Journal of Lexicography, 16:235–250. find
  6. Charles J. Fillmore. 1976. Frame semantics and the nature of language. Annals of the New York Academy of Sciences: Conference on the Origin and Development of Language and Speech, 280:20–32. find
  7. Hagen Fürstenau. 2008. Enriching frame semantic resources with dependency graphs. In Proceedings of the 6th Language Resources and Evaluation Conference, pages 1478–1484, Marrakech, Morocco. find
  8. Daniel Gildea and Daniel Jurafsky. 2002. Automatic labeling of semantic roles. Computational Linguistics, 28(3):245–288. find our card
  9. K. Hacioglu. 2004. Semantic role labeling using dependency trees. In Proceedings of the 20th international conference on Computational Linguistics, page 1273. Association for Computational Linguistics. find
  10. Richard Johansson and Pierre Nugues. 2007. Lth: semantic structure extraction using nonprojective dependency trees. In Proceedings of of SemEval. find
  11. Dan Klein and Christopher D. Manning. 2003. Accurate unlexicalized parsing. In Proceedings of the 41st Meeting of the Association for Computational Linguistics, pages 423–430. find our card
  12. A. Moschitti, D. Pighin, and R. Basili. 2008. Tree kernels for semantic role labeling. Computational Linguistics, 34(2):193–224. find
  13. Martha Palmer, Daniel Gildea, and Paul Kingsbury. 2005. The Proposition Bank: An annotated corpus of semantic roles. Computational Linguistics, 31(1):71–106. find our card
  14. Owen Rambow, Bonnie Dorr, Karin Kipper, Ivona Kučerová, and Martha Palmer. 2003. Automatically deriving tectogrammatical labels from other resources: A comparison of semantic labels across frameworks. The Prague Bulletin of Mathematical Linguistics, (79– 80):23–36. find
  15. Mihai Surdeanu, Richard Johansson, Adam Meyers, Lluı́s Màrquez, and Joakim Nivre. 2008. find
  16. The CoNLL 2008 shared task on joint parsing of syntactic and semantic dependencies. In Proceedings of CoNLL 2008, pages 159– 177, Manchester, England. find our card

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/papers/bauer-2012-dependency-parsed-framenet-corpus/