Elamur Research⌕
☰
home / papers / 2015 / A Transition-based Algorithm for AMR Parsing

A Transition-based Algorithm for AMR Parsing

Summary

The paper addresses automatic parsing of sentences into Abstract Meaning Representation (AMR). AMR is a rooted directed graph: its nodes denote concepts, its edges denote semantic relations, and a single concept can participate in several relations. Rather than building the graph directly with complex algorithms, the authors propose a two-stage scheme. First, an ordinary dependency parser builds a syntactic tree. Then a new transition-based algorithm transforms this tree into an AMR graph step by step. For training, the words of a sentence are aligned with AMR concepts using the aligner from JAMR. The result is stored as a "span graph", in which each node corresponds to a contiguous span of text. Eight linguistically motivated actions make it possible to predict nodes, edges and their labels simultaneously. The main advantage of the approach is that the syntactic parser can be trained on a much larger treebank than the small AMR corpus. On the test data, the system achieved a Smatch F-score of 0.63, 5 points above the previous best result. Thus, the syntactic tree proved to be a good starting point for semantic parsing, and the method itself runs fast in practice.

Key points

  • Eight transition actions: NEXT-EDGE, SWAP, REATTACH, REPLACE-HEAD, REENTRANCE, MERGE, NEXT-NODE, DELETE-NODE. The tree is processed bottom-up, in post-order.
  • The input trees are built by Stanford CoreNLP v3.3.1 (the wsjPCFG model, trained on sections 02–21 of the WSJ treebank). The data is the AMR Annotation Corpus, more than 10 thousand sentences.
  • Smatch on the test set: precision 0.64, recall 0.62, F 0.63; JAMR achieves F 0.58. In variants with gold-standard labels substituted, F rises to 0.71.
  • When the actions are chosen by an oracle, the gold graphs on the training data can be reproduced with F ≈ 99%. Hence, the action set is almost always sufficient for this transformation.
  • The worst-case complexity is O(n²), but in practice the parser runs in nearly linear time. The actions most often in error are REATTACH, REENTRANCE and SWAP: these are associated with prepositional phrase attachment (PP-attachment) errors and with coordinate constructions.

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 · Machine translation

Topics Semantic parsing & meaning representation

Cite

Cite as Chuan Wang, Nianwen Xue and Sameer Pradhan. 2015. A Transition-based Algorithm for AMR Parsing. Proceedings of the 2015 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.3115/v1/n15-1040

@inproceedings{wang2015transition,
  title = {{A Transition-based Algorithm for AMR Parsing}},
  author = {Chuan Wang, Nianwen Xue and Sameer Pradhan},
  booktitle = {Proceedings of the 2015 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies},
  year = {2015},
  doi = {10.3115/v1/n15-1040},
  url = {https://doi.org/10.3115/v1/n15-1040}
}

References

16 references, 11 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Laura Banarescu, Claire Bonial, Shu Cai, Madalina Georgescu, Kira Griffitt, Ulf Hermjakob, Kevin Knight, Philipp Koehn, Martha Palmer, and Nathan Schneider, 2013. Abstract Meaning Representation for Sembanking. In Proceedings of the 7th Linguistic Annotation Workshop and Interoperability with Discourse, pages 178–186. Association for Computational Linguistics. our card
  2. Bernd Bohnet and Joakim Nivre. 2012. A transition-based system for joint part-of-speech tagging and labeled non-projective dependency parsing. In Proceedings of the 2012 Joint Conference on Empirical Methods in Natural Language Processing and Computational Natural Language Learning, pages 1455–1465. Association for Computational Linguistics. our card
  3. Shu Cai and Kevin Knight. 2013. Smatch: an evaluation metric for semantic feature structures. In Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers), pages 748–752. Association for Computational Linguistics. find
  4. David Chiang, Jacob Andreas, Daniel Bauer, Karl Moritz Hermann, Bevan Jones, and Kevin Knight. 2013. Parsing graphs with hyperedge replacement grammars. In Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 924–932, Sofia, Bulgaria, August. Association for Computational Linguistics. find
  5. Michael Collins and Brian Roark. 2004. Incremental parsing with the perceptron algorithm. In Proceedings of the 42nd Annual Meeting on Association for Computational Linguistics, page 111. Association for Computational Linguistics. our card
  6. Michael Collins. 2002. Discriminative training methods for hidden markov models: Theory and experiments with perceptron algorithms. In Proceedings of the ACL-02 conference on Empirical methods in natural language processing-Volume 10, pages 1–8. Association for Computational Linguistics. our card
  7. Jeffrey Flanigan, Sam Thomson, Jaime Carbonell, Chris Dyer, and Noah A. Smith. 2014. A discriminative graph-based parser for the abstract meaning representation. In Proceedings of the 52nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 1426–1436, Baltimore, Maryland, June. Association for Computational Linguistics. our card
  8. Dan Klein and Christopher D Manning. 2003. Accurate unlexicalized parsing. In Proceedings of the 41st Annual Meeting on Association for Computational Linguistics-Volume 1, pages 423–430. Association for Computational Linguistics. our card
  9. Christopher D. Manning, Mihai Surdeanu, John Bauer, Jenny Finkel, Steven J. Bethard, and David McClosky. 2014. The Stanford CoreNLP natural language processing toolkit. In Proceedings of 52nd Annual Meeting of the Association for Computational Linguistics: System Demonstrations, pages 55–60. our card
  10. Joakim Nivre. 2003. An efficient algorithm for projective dependency parsing. In Proceedings of the 8th International Workshop on Parsing Technologies (IWPT). Citeseer. our card
  11. Joakim Nivre. 2007. Incremental non-projective dependency parsing. In Human Language Technologies 2007: The Conference of the North American Chapter of the Association for Computational Linguistics; Proceedings of the Main Conference, pages 396–403. Association for Computational Linguistics. find
  12. Joakim Nivre. 2008. Algorithms for deterministic incremental dependency parsing. Computational Linguistics, 34(4):513–553. our card
  13. Joakim Nivre. 2009. Non-projective dependency parsing in expected linear time. In Proceedings of the Joint Conference of the 47th Annual Meeting of the ACL and the 4th International Joint Conference on Natural Language Processing of the AFNLP: Volume 1-Volume 1, pages 351–359. Association for Computational Linguistics. our card
  14. Kenji Sagae and Jun’ichi Tsujii. 2008. Shift-reduce dependency dag parsing. In Proceedings of the 22nd International Conference on Computational Linguistics-Volume 1, pages 753–760. Association for Computational Linguistics. find
  15. Francesco Sartorio, Giorgio Satta, and Joakim Nivre. 2013. A transition-based dependency parser using a dynamic parsing strategy. In Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 135–144. Association for Computational Linguistics. find
  16. Hiroyasu Yamada and Yuji Matsumoto. 2003. Statistical dependency analysis with support vector machines. In Proceedings of IWPT, volume 3. our card

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/wang-2015-transition-based-algorithm-amr-parsing/