Reassessing Graph Linearization for Sequence-to-sequence AMR Parsing: On the Advantages and Limitations of Triple-Based Encoding

TL;DR

This paper evaluates a triple-based linearization method for AMR parsing, highlighting its advantages over Penman encoding in representing complex graphs, but also noting current limitations in efficiency.

Contribution

It introduces a novel triple-based linearization approach for AMR graphs, addressing issues of node proximity and re-entrancy handling in sequence-to-sequence models.

Findings

Triple-based encoding better captures graph structure.

Penman encoding remains more concise and explicit.

Room for improvement in triple encoding efficiency.

Abstract

Sequence-to-sequence models are widely used to train Abstract Meaning Representation (Banarescu et al., 2013, AMR) parsers. To train such models, AMR graphs have to be linearized into a one-line text format. While Penman encoding is typically used for this purpose, we argue that it has limitations: (1) for deep graphs, some closely related nodes are located far apart in the linearized text (2) Penman's tree-based encoding necessitates inverse roles to handle node re-entrancy, doubling the number of relation types to predict. To address these issues, we propose a triple-based linearization method and compare its efficiency with Penman linearization. Although triples are well suited to represent a graph, our results suggest room for improvement in triple encoding to better compete with Penman's concise and explicit representation of a nested graph structure.