Probabilistic, Structure-Aware Algorithms for Improved Variety, Accuracy, and Coverage of AMR Alignments

Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing, pages 3310–3321 August 1–6, 2021. ©2021 Association for Computational Linguistics 3310 Probabilistic, Structure-Aware Algorithms for Improved Variety, Accuracy, and Coverage of AMR Alignments Austin Blodgett Nathan Schneider Georgetown University {ajb341, nathan.schneider}@georgetown.edu Abstract We present algorithms for aligning compo- nents of Abstract Meaning Representation (AMR) graphs to spans in English sentences. We leverage unsupervised learning in combi- nation with heuristics, taking the best of both worlds from previous AMR aligners. Our un- supervised models, however, are more sensi- tive to graph substructures, without requiring a separate syntactic parse. Our approach covers a wider variety of AMR substructures than pre- viously considered, achieves higher coverage of nodes and edges, and does so with higher accuracy. We will release our LEAMR datasets and aligner for use in research on AMR pars- ing, generation, and evaluation. 1 Introduction Research with the Abstract Meaning Represen- tation (AMR; Banarescu et al., 2013), a broad- coverage semantic annotation framework in which sentences are paired with directed acyclic graphs, must contend with the lack of gold-standard align- ments between words and semantic units in the English data. A variety of rule-based and statistical algorithms have sought to fill this void, with im- provements in alignment accuracy often translating into improvements in AMR parsing accuracy (Pour- damghani et al., 2014; Naseem et al., 2019; Liu et al., 2018). Yet current alignment algorithms still suffer from limited coverage and less-than-ideal accuracy, constraining the design and accuracy of parsing algorithms. Where parsers use latent align- ments (e.g., Lyu and Titov, 2018; Cai and Lam, 2020), explicit alignments can still facilitate evalu- ation and error analysis. Moreover, AMR-to-text generation research and applications using AMR stand to benefit from accurate, human-interpretable alignments. We present Linguistically Enriched AMR (LEAMR) alignment, which achieves full graph cov- erage via four distinct types of aligned structures: subgraphs, relations, reentrancies, and duplicate subgraphs arising from ellipsis. This formulation lends itself to unsupervised learning of alignment models. Advantages of our algorithm and released alignments include: (1) much improved coverage over previous datasets, (2) increased variety of the substructures aligned, including alignments for all relations, and alignments for diagnosing reentran- cies, (3) alignments are made between spans and connected substructures of an AMR, (4) broader identification of spans including named entities and verbal and prepositional multiword expressions. Contributions are as follows: • A novel all-inclusive formulation of AMR alignment in terms of mappings between spans and connected subgraphs, including spans aligned to multiple subgraphs; map- pings between spans and inter-subgraph edges; and characterization of reentrancies. Together these alignments fully cover the nodes and edges of the AMR graph (§3). • An algorithm combining rules and EM to align English sentences to AMRs without su- pervision (§5), achieving higher coverage and quality than existing AMR aligners (§7). • A corpus with automatic alignments for LDC2020 and Little Prince data as well as a few hundred manually annotated sentences for …

Content truncated to 500 words