Elamur Research
☰
home / papers / 2017 / An Incremental Parser for Abstract Meaning Representation

An Incremental Parser for Abstract Meaning Representation

Written for the Research library (claude-opus-5-5) from the paper text. Original — aclanthology.org/E17-1051.pdf

Summary

The paper tackles parsing sentences into Abstract Meaning Representation (AMR), a graph-based semantic formalism that combines named entities, semantic roles, word senses and coreference in one structure. The authors adapt transition-based dependency parsing to this task. Their system, AMR-EAGER, builds on ARC-EAGER (Nivre, 2004) and reads a sentence once, left to right, in worst-case linear time. To make this work for AMR, they map word tokens to graph fragments, allow nodes with several parents (reentrancy), and use a modified Reduce transition that can add edges between sibling nodes. This last change captures control-verb patterns and some non-projective structures. A feed-forward neural network chooses each transition greedily. The authors also argue that a single Smatch score hides important differences between parsers, so they release a suite of metrics for individual subtasks. On LDC2015E86, the parser is competitive but does not lead on overall Smatch. It is strongest on named entities and negation. The work matters because it brings AMR closer to fast, incremental dependency-parsing methods and offers a more detailed way to compare parsers.

Key points

  • In 200 manually aligned AMRs, only 6% of edges are non-projective, yet 51% of graphs contain at least one such edge. Reentrancy appears in 93% of graphs.
  • On LDC2015E86, AMR-EAGER reaches a Smatch of 64, compared with 67 for JAMR (2016), 63 for CAMR (2015) and 58 for JAMR (2014).
  • The parser is best on Named Ent. (83) and Negations (48). It ties for the top Unlabeled (69) and Concepts (83) scores and comes within one point of the best Reentrancy score (41 vs 42).
  • The evaluation suite scores several subtasks separately: unlabeled graphs, no WSD, NP-only, reentrancy, concepts, named entities, wikification, negation, and SRL (Smatch over :ARG edges).
  • The code is public: the parser at github.com/mdtux89/amr-eager and the metrics at github.com/mdtux89/amr-evaluation, plus an online demo. By contrast, CAMR is quadratic in the worst case, and parts of JAMR run in more than linear time.

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: Named entity recognition · Dependency parsing · Semantic role labeling

Cited by

116 papers of our library cite it (of the 140673 whose bibliographies we read); 9 of them have a page here, 107 not yet.

Cite

The paper itself, not this page. BibTeX from the library's record of the work. arXiv 1608.06111

@inproceedings{damonte2017incremental,
  title = {{An Incremental Parser for Abstract Meaning Representation}},
  author = {Damonte, Marco and Cohen, Shay B. and Satta, Giorgio},
  booktitle = {Proceedings of the 15th Conference of the European Chapter of the Association for Computational Linguistics: Volume 1, Long Papers},
  year = {2017},
  eprint = {1608.06111},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/1608.06111}
}

References

24 references, 4 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Xiaochang Peng, Linfeng Song, and Daniel Gildea. 2015. A synchronous hyperedge replacement grammar based approach for AMR parsing. Proceedings of CoNLL. find
  2. Omri Abend and Ari Rappoport. 2013. Universal conceptual cognitive annotation (UCCA). In Proceedings of ACL. find our card
  3. Michael Pust, Ulf Hermjakob, Kevin Knight, Daniel Marcu, and Jonathan May. 2015. Using syntax-based machine translation to parse english into abstract meaning representation. arXiv preprint arXiv:1504.06665. ↗
  4. Yoav Artzi, Kenton Lee, and Luke Zettlemoyer. 2015. Broad-coverage CCG semantic parsing with AMR. Proceedings of EMNLP. find
  5. 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. Proceedings of Linguistic Annotation Workshop. find our card
  6. Sudh Rao, Yogarshi Vyas, Hal Daume III, and Philip Resnik. 2015. Parser for abstract meaning representation using learning to search. arXiv:1510.07586. ↗
  7. Corentin Ribeyre, Éric Villemonte de La Clergerie, and Djamé Seddah. 2015. Because syntax does matter: Improving predicate-argument structures parsing using syntactic features. In Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. find
  8. Shu Cai and Kevin Knight. 2013. Smatch: an evaluation metric for semantic feature structures. Proceedings of ACL. find
  9. Danqi Chen and Christopher D Manning. 2014. A fast and accurate dependency parser using neural networks. In Proceesings of EMNLP. find our card
  10. Kenji Sagae and Jun’ichi Tsujii. 2008. Shift-reduce dependency DAG parsing. Proceedings of COL-ING. find
  11. Yantao Du, Fan Zhang, Weiwei Sun, and Xiaojun Wan. 2014. Peking: Profiling syntactic tree parsing techniques for semantic graph parsing. In Proceedings of the 8th International Workshop on Semantic Evaluation (SemEval-2014), pages 459–464. find
  12. Yuichiro Sawai, Hiroyuki Shindo, and Yuji Matsumoto. 2015. Semantic structure analysis of noun phrases using abstract meaning representation. Proceedings of ACL. find
  13. Jeffrey Flanigan, Sam Thomson, Jaime G Carbonell, Chris Dyer, and Noah A Smith. 2014. A discriminative graph-based parser for the abstract meaning representation. Proceedings of ACL. find
  14. Mark Steedman. 1996. Surface Structure and Interpretation. The MIT Press. find
  15. Mark Steedman. 2000. The Syntactic Process. The MIT Press. find
  16. Jeffrey Flanigan, Chris Dyer, Noah A Smith, and Jaime Carbonell. 2016. CMU at SemEval-2016 task 8: Graph-based AMR parsing with infinite ramp loss. Proceedings of SemEval, pages 1202–1206. find
  17. Lucy Vanderwende, Arul Menezes, and Chris Quirk. 2015. An AMR parser for english, french, german, spanish and japanese and a new AMR-annotated corpus. Proceedings of NAACL-HLT. find
  18. James Goodman, Andreas Vlachos, and Jason Naradowsky. 2016. Noise reduction and targeted exploration in imitation learning for abstract meaning representation parsing. Proceedings of ACL. find
  19. Chuan Wang, Nianwen Xue, and Sameer Pradhan. 2015a. Boosting transition-based AMR parsing with refined actions and auxiliary analyzers. Proceedings of ACL. find our card
  20. Marco Kuhlmann and Peter Jonsson. 2015. Parsing to noncrossing dependency graphs. Transactions of the Association for Computational Linguistics, pages 559–570. find
  21. Chuan Wang, Nianwen Xue, and Sameer Pradhan. 2015b. A transition-based algorithm for AMR parsing. Proceedings of NAACL. find
  22. Chuan Wang, Sameer Pradhan, Nianwen Xue, Xiaoman Pan, and Heng Ji. 2016. CAMR at SemEval-2016 task 8: An extended transition-based AMR parser. Proceedings of SemEval. find
  23. Yue Zhang and Joakim Nivre. 2011. Transition-based dependency parsing with rich non-local features. Proceedings of ACL. find
  24. Junsheng Zhou, Feiyu Xu, Hans Uszkoreit, Weiguang QU, Ran Li, and Yanhui Gu. 2016. AMR parsing with an incremental joint model. In Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing. find