AMR Dependency Parsing with a Typed Semantic Algebra

Abstract

We present a semantic parser for Abstract Meaning Representations which learns to parse strings into tree representations of the compositional structure of an AMR graph. This allows us to use standard neural techniques for supertagging and dependency tree parsing, constrained by a linguistically principled type system. We present two approximative decoding algorithms, which achieve state-of-the-art accuracy and outperform strong baselines.

This paper will be presented at ACL 2018 (see https://acl2018.org/programme/papers/)

References in corpus (3)

• Grammar as a Foreign Language
• Neural Semantic Parsing by Character-based Translation: Experiments with Abstract Meaning Representations
• Universal Semantic Parsing

Cited by in corpus (13)

• Semantic Neural Machine Translation using AMR
• AMR Parsing as Sequence-to-Graph Transduction
• AMR Parsing via Graph-Sequence Iterative Inference
• AMR Parsing with Action-Pointer Transformer
• Core Semantic First: A Top-down Approach for AMR Parsing
• Tackling Graphical NLP problems with Graph Recurrent Networks
• A Comparative Analysis of Knowledge-Intensive and Data-Intensive Semantic Parsers
• SemBleu: A Robust Metric for AMR Parsing Evaluation
• Automatic Accuracy Prediction for AMR Parsing
• The First Shared Task on Discourse Representation Structure Parsing
• Polynomial Graph Parsing with Non-Structural Reentrancies
• A Differentiable Relaxation of Graph Segmentation and Alignment for AMR Parsing
• Semantic Graph Parsing with Recurrent Neural Network DAG Grammars