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

Source: [Probabilistic, Structure-Aware Algorithms for Improved Variety, Accuracy, and Coverage of AMR Alignments](<https://www.sota2.com/research/paper/68fa6a201e209131cea26222>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-25T23:16:37.291Z

Authors: Austin Blodgett, Nathan Schneider

Published: 2021-06-10T18:46:32.000Z

## Abstract

We present algorithms for aligning components of Abstract Meaning Representation (AMR) graphs to spans in English sentences. We leverage unsupervised learning in combination with heuristics, taking the best of both worlds from previous AMR aligners. Our unsupervised models, however, are more sensitive to graph substructures, without requiring a separate syntactic parse. Our approach covers a wider variety of AMR substructures than previously 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 parsing, generation, and evaluation.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Reentrancy Alignment on LEAMR 1.0 (test)](<https://www.sota2.com/research/sota/reentrancy-alignment-on-leamr-1-0-test.md>): Reentrancy Alignment; LEAMR 1.0 (test); Exact Alignment Precision: 55.75.

- [Subgraph Alignment on LEAMR 1.0 (test)](<https://www.sota2.com/research/sota/subgraph-alignment-on-leamr-1-0-test.md>): Subgraph Alignment; LEAMR 1.0 (test); Exact Alignment Precision: 93.91.

- [Relation Alignment on LEAMR 1.0 (test)](<https://www.sota2.com/research/sota/relation-alignment-on-leamr-1-0-test.md>): Relation Alignment; LEAMR 1.0 (test); Exact Alignment P: 0.8567.

- [Inter-lingual Subgraph Identification on ISI German](<https://www.sota2.com/research/sota/inter-lingual-subgraph-identification-on-isi-german.md>): Inter-lingual Subgraph Identification; ISI German; Precision: 8.4.

- [Inter-lingual Subgraph Identification on ISI Spanish (ES)](<https://www.sota2.com/research/sota/inter-lingual-subgraph-identification-on-isi-spanish-es.md>): Inter-lingual Subgraph Identification; ISI Spanish (ES); Precision: 8.1.

- [Inter-lingual Subgraph Identification on ISI Italian (IT)](<https://www.sota2.com/research/sota/inter-lingual-subgraph-identification-on-isi-italian-it.md>): Inter-lingual Subgraph Identification; ISI Italian (IT); Precision: 9.

- [Inter-lingual Subgraph Identification on ISI English](<https://www.sota2.com/research/sota/inter-lingual-subgraph-identification-on-isi-english.md>): Inter-lingual Subgraph Identification; ISI English; Precision: 85.9.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2106.06002>)

[PDF](<https://arxiv.org/pdf/2106.06002.pdf>)