DocAMR: Multi-Sentence AMR Representation and Evaluation

What it is about

AMR was built for single sentences, so a document needs one graph that joins sentences through their coreference links. The earlier proposal merged all coreferent nodes into one node. The paper shows that this loses or distorts meaning: a negative polarity can move onto a merged node and reverse a sentence, and merged predicates can lose their arguments. The paper proposes DocAMR, which keeps coreference as links between the sentence graphs rather than merging them. It also makes Smatch fast enough for document graphs and adds a coreference sub-score. On the gold double-annotated development set, Smatch is 93.3 with no merge, 92.8 when only named entities are merged, 87.6 for DocAMR, and 82.4 for full merge. Run times are 66, 66, 90 and 153 seconds. A pipeline of a top AMR parser and coreference systems gives a baseline that beats the earlier best system by a large margin.

Ideas

• I1. Typed coreference links in the meta-world. Do not merge meta tokens that refer to the same participant across verses. Keep them as typed links. Count how many merges would flip a negative or change a predicate's arguments on the Bible pairs, the paper's own failure cases.
• I2. A coreference sub-score per author. Pronoun reference is where thou and you differ between translations. Add a reference class to measure.rs and report it separately, so an author delta on pronouns is visible.
• I3. Cost of the graph. Time the matching with and without the cross-verse links, and keep the outputs identical across two runs, as the project requires.

For Mova

The merge failure is the same kind of error that open problem 2 describes: merges that look right but lose meaning. Keeping coreference as links, rather than merging, fits the meta-world's concept-plus-link design. The pronoun class in I2 bears on open problem 5 and on the author deltas in profile.rs.

Topics: Coreference resolution · Machine translation

Cite

The paper itself, not this card. BibTeX from the library's record of the work. DOI 10.18653/v1/2022.naacl-main.256

@inproceedings{naseem2022docamr,
  title = {{DocAMR: Multi-Sentence AMR Representation and Evaluation}},
  author = {Naseem, Tahira and Blodgett, Austin and Kumaravel, Sadhana and O'Gorman, Tim and Lee, Young-Suk and Flanigan, Jeffrey and Astudillo, Ramón and Florian, Radu and Roukos, Salim and Schneider, Nathan},
  booktitle = {Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies},
  year = {2022},
  doi = {10.18653/v1/2022.naacl-main.256},
  url = {https://doi.org/10.18653/v1/2022.naacl-main.256}
}

References

13 references, 1 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Tatiana Anikina, Alexander Koller, and Michael Roth. 2020. Predicting coreference in Abstract Meaning Representations. In Proceedings of the Third Workshop on Computational Models of Reference, Anaphora and Coreference, pages 33–38, Barcelona, Spain (online). Association for Computational Linguistics. find
2. Xuefeng Bai, Yulong Chen, Linfeng Song, and Yue Zhang. 2021. Semantic representation for dialogue modeling. In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers), pages 4430–4445, Online. Association for Computational Linguistics. find
3. 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, Sofia, Bulgaria. Association for Computational Linguistics. find our card
4. 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, Sofia, Bulgaria. Association for Computational Linguistics. find
5. Kevin Clark and Christopher D. Manning. 2016. Improving coreference resolution by learning entity-level distributed representations. In Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 643–653, Berlin, Germany. Association for Computational Linguistics. find
6. Marco Damonte, Shay B. Cohen, and Giorgio Satta. 2017. An incremental parser for Abstract Meaning Representation. In Proceedings of the 15th Conference of the European Chapter of the Association for Computational Linguistics: Volume 1, Long Papers, pages 536–546, Valencia, Spain. Association for Computational Linguistics. find
7. Qiankun Fu, Linfeng Song, Wenyu Du, and Yue Zhang. 2021. End-to-end AMR coreference resolution. In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers), pages 4204– 4214, Online. Association for Computational Linguistics. find
8. Mandar Joshi, Danqi Chen, Yinhan Liu, Daniel S. Weld, Luke Zettlemoyer, and Omer Levy. 2020. Span-BERT: Improving pre-training by representing and predicting spans. Transactions of the Association for Computational Linguistics, 8:64–77. find
9. Fei-Tzin Lee, Chris Kedzie, Nakul Verma, and Kathleen McKeown. 2021. An analysis of document graph construction method for AMR summarization. arXiv preprint: arXiv:2111.13993v1. ↗
10. Matthias Lindemann, Jonas Groschwitz, and Alexander Koller. 2019. Compositional semantic parsing across graphbanks. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, pages 4576–4585, Florence, Italy. Association for Computational Linguistics. find
11. Fei Liu, Jeffrey Flanigan, Sam Thomson, Norman Sadeh, and Noah A. Smith. 2015. Toward abstractive summarization using semantic representations. In Proceedings of the 2015 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 1077–1086, Denver, Colorado. Association for Computational Linguistics. find
12. Tim O’Gorman, Michael Regan, Kira Griffitt, Ulf Hermjakob, Kevin Knight, and Martha Palmer. 2018. AMR beyond the sentence: the multi-sentence AMR corpus. In Proceedings of the 27th International Conference on Computational Linguistics, pages 3693– 3702, Santa Fe, New Mexico, USA. Association for Computational Linguistics. find
13. Jiawei Zhou, Tahira Naseem, Ramón Fernandez Astudillo, Young-Suk Lee, Radu Florian, and Salim Roukos. 2021. Structure-aware fine-tuning of sequence-to-sequence transformers for transition-based AMR parsing. In Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, pages 6279–6290, Online and Punta Cana, Dominican Republic. Association for Computational Linguistics. find

This page in Markdown → https://md.mova.elamur.ai/free/en/docs/papers/naseem-2022-docamr-multi-sentence-amr-representation/