A Partially Rule-Based Approach to AMR Generation

What it is about

The paper generates English from AMR with mostly hand-written rules and only light statistics. Each AMR concept gets a set of candidate realisations: frame concepts become verbal, adjectival and nominal forms (with -ing, -ed, -tion, -ly variants), and special concepts such as named entities and role relations get fixed rules. Relations become prepositions or affixes. Candidates are combined by cube pruning over a 5-gram language model, and the children of a node are ordered by a pair-order model that counts how often one relation precedes another (for example, ARG0 comes before its head 77% of the time). Nodes with more than five children fall back to three baseline orders. The author argues that BLEU is a poor metric here and reports that METEOR and TER track human judgements better, though the system scores below the statistical systems on BLEU.

Ideas

• I1. Port the pair-order model into en/src/lin.rs as a count table: for each pair of relations under one concept, count the order seen in aligned data, add one, and score each permutation by its pairs. Compare with the side tables on crossings first.
• I2. Bounded search: score all orders of a node with up to five children (at most 720), and fall back to a small set of baseline orders above that. This is deterministic and keeps the cost known in advance.
• I3. Form candidates from the concept name, as the paper does, and let counts from the corpus prune the implausible ones. For Mova this is a candidate list in en/src/morph.rs, with pruning by counts instead of a language model.
• I4. Add METEOR and TER to the translator's normal mode next to exact match and chrF. Check that the ranking of systems on dev is the same across two identical runs before trusting either metric.

For Mova

The order model fits en/src/lin.rs, the candidate list fits en/src/morph.rs, and the metrics fit the normal mode of tools/translate/measure.rs. The hand-written realisation rules are the part of the paper that shows the control Mova wants: each rule is readable and its effect can be traced.

Topics: Machine translation

Cite

The paper itself, not this card. BibTeX from the library's record of the work. DOI 10.18653/v1/n19-3009

@inproceedings{manning2019partially,
  title = {{A Partially Rule-Based Approach to AMR Generation}},
  author = {Manning, Emma},
  booktitle = {Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Student Research Workshop},
  year = {2019},
  doi = {10.18653/v1/n19-3009},
  url = {https://doi.org/10.18653/v1/n19-3009}
}

References

21 references, 9 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 & Interoperability with Discourse, pages 178–186, Sofia, Bulgaria. find our card
2. Philipp Koehn, Hieu Hoang, Alexandra Birch, Chris Callison-Burch, Marcello Federico, Nicola Bertoldi Itc-Irst, Brooke Cowan, Wade Shen, Christine Moran Mit, Richard Zens, Rwth Aachen, Chris Dyer, Alexandra Constantin, Williams College, and Evan Herbst Cornell. 2007. Moses: Open Source Toolkit for Statistical Machine Translation. In Proceedings of the 45th Annual Meeting of the Association of Computational Linguistics, pages 177–180, Prague, Czech Republic. find
3. Steven Bird, Edward Loper, and Ewan Klein. 2009. Natural Language Processing with Python. O’Reilly Media Inc. find
4. Thiago Castro Ferreira, Iacer Calixto, Sander Wubben, and Emiel Krahmer. 2017. Linguistic realisation as machine translation: Comparing different MT models for AMR-to-text generation. In Proceedings of the 10th International Conference on Natural Language Generation., pages 1–10, Santiago de Compostela, Spain. Association for Computational Linguistics. find our card
5. Ioannis Konstas, Srinivasan Iyer, Mark Yatskar, Yejin Choi, Luke Zettlemoyer, and Paul G Allen. 2017. Neural AMR: Sequence-to-Sequence Models for Parsing and Generation. In Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics, pages 146–157, Vancouver, Canada. find our card
6. Leshem Choshen and Omri Abend. 2018. Reference-less Measure of Faithfulness for Grammatical Error Correction. In Proceedings of NAACL-HLT 2018, pages 124–129, New Orleans, Louisiana. Association for Computational Linguistics. find
7. Gerasimos Lampouras and Andreas Vlachos. 2017. Sheffield at SemEval-2017 Task 9: Transition-based language generation from AMR. In Proceedings of the 11th International Workshop on Semantic Evaluations (SemEval-2017), pages 586–591, Vancouver, Canada. Association for Computational Linguistics. find
8. Michael Denkowski and Alon Lavie. 2014. Meteor Universal: Language Specific Translation Evaluation for Any Target Language. In Proceedings of the EACL 2014 Workshop on Statistical Machine Translation, pages 376–380, Baltimore, Maryland. find
9. Alon Lavie and Abhaya Agarwal. 2007. Meteor: An Automatic Metric for MT Evaluation with High Levels of Correlation with Human Judgments. In Proceedings of the Second Workshop on Statistical Machine Translation, pages 228–231, Prague, Czech Republic. Association for Computational Linguistics. Austin, TX. Association for Computational Linguistics. find
10. Jonathan May and Jay Priyadarshi. 2017. SemEval-2017 Task 9: Abstract Meaning Representation Parsing and Generation. In Proceedings of the 11th International Workshop on Semantic Evaluations, pages 536–545, Vancouver, Canada. find
11. Linfeng Song, Yue Zhang, Zhiguo Wang, and Daniel Gildea. 2018. A Graph-to-Sequence Model for AMR-to-Text Generation. In Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Long Papers), pages 1616– 1626, Melbourne, Australia. Association for Computational Linguistics. find
12. Simon Mille, Roberto Carlini, Alicia Burga, and Leo Wanner. 2017. FORGe at SemEval-2017 Task 9: Deep sentence generation based on a sequence of graph transducers. In Proceedings of the 11th International Workshop on Semantic Evaluations (SemEval-2017), pages 920–923, Vancouver, Canada. Association for Computational Linguistics. find
13. Ida Szubert, Adam Lopez, and Nathan Schneider. 2018. A Structured Syntax-Semantics Interface for English-AMR Alignment. In Proceedings of NAACL-HLT 2018, pages 1169–1180, New Orleans, Louisiana. Association for Computational Linguistics. find our card
14. Courtney Napoles, Keisuke Sakaguchi, and Joel Tetreault. 2016. There’s No Comparison: Reference-less Evaluation Metrics in Grammatical Error Correction. In Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing, pages 2109–2115, Austin, TX. Association for Computational Linguistics. find
15. Kishore Papineni, Salim Roukos, Todd Ward, and Wei-Jing Zhu. 2002. BLEU: a Method for Automatic Evaluation of Machine Translation. In Proceedings of the 40th Annual Meeting of the Association for Computational Linguistics, pages 311–318, Philadelphia, Pennsylvania. find our card
16. Nima Pourdamghani, Kevin Knight, and Ulf Hermjakob. 2016. Generating English from Abstract Meaning Representations. In Proceedings of The 9th International Natural Language Generation conference, pages 21–25, Edinburgh, UK. find our card
17. Beatrice Santorini. 1990. Part-of-Speech Tagging Guidelines for the Penn Treebank Project (3rd Revision). Technical report, University of Pennsylvania. find our card
18. Timo Schick. 2017. Transition-Based Generation from Abstract Meaning Representations. Ph.D. thesis, Technische Universität Dresden. find
19. Matthew Snover, Bonnie Dorr, Richard Schwartz, Linnea Micciulla, and John Makhoul. 2006. A Study of Translation Edit Rate with Targeted Human Annotation. In Proceedings of the 7th Conference of the Association for Machine Translation in the Americas, pages 223–231, Cambridge. find our card
20. Linfeng Song, Xiaochang Peng, Yue Zhang, Zhiguo Wang, and Daniel Gildea. 2017. AMR-to-text Generation with Synchronous Node Replacement Grammar. In Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Short Papers), pages 7–13, Vancouver, Canada. Association for Computational Linguistics. find our card
21. Linfeng Song, Yue Zhang, Xiaochang Peng, Zhiguo Wang, and Daniel Gildea. 2016. AMR-to-text generation as a Traveling Salesman Problem. In Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing, pages 2084–2089, find

This page in Markdown → https://md.mova.elamur.ai/free/en/docs/papers/manning-2019-partially-rule-based-approach-amr/