AMR Parsing as Graph Prediction with Latent Alignment

Chunchuan Lyu, Ivan Titov
5/14/2018
cs.CL

Official code: ChunchuanLv/AMR_AS_GRAPH_PREDICTION (54 stars, last commit 2021-01-09).

Abstract

Abstract meaning representations (AMRs) are broad-coverage sentence-level semantic representations. AMRs represent sentences as rooted labeled directed acyclic graphs. AMR parsing is challenging partly due to the lack of annotated alignments between nodes in the graphs and words in the corresponding sentences. We introduce a neural parser which treats alignments as latent variables within a joint probabilistic model of concepts, relations and alignments. As exact inference requires marginalizing over alignments and is infeasible, we use the variational auto-encoding framework and a continuous relaxation of the discrete alignments. We show that joint modeling is preferable to using a pipeline of align and parse. The parser achieves the best reported results on the standard benchmark (74.4% on LDC2016E25).

View on arXivView PDF

Code Implementations(1)

ChunchuanLv/AMR_AS_GRAPH_PREDICTIONOfficial100%
5416Java, PythonMay 9, 20185 years ago

Similar Papers

Coarse-to-Fine Decoding for Neural Semantic Parsing

83% match1 repo1682018

Word Alignment-Based Evaluation of Uniform Meaning Representations

83% match2026

Incorporating Graph Information in Transformer-based AMR Parsing

82% match1 repo82023

Rethinking Relation Classification with Graph Meaning Representations

82% match2023

ERAlign: Energy-based Representation Alignment of GNNs and LLMs on Text-attributed Graphs

82% match2026

Joint Representations of Text and Knowledge Graphs for Retrieval and Evaluation

82% match2023

Cite this paper

@article{lyu2018parsing,
  title  = {AMR Parsing as Graph Prediction with Latent Alignment},
  author = {Chunchuan Lyu and Ivan Titov},
  year   = {2018},
  eprint = {1805.05286},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/abs/1805.05286}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/1805.05286.svg)](https://codeofpaper.com/paper/1805.05286)

Discussion