Hierarchical Bracketing Encodings for Dependency Parsing as Tagging

TL;DR

This paper introduces a new hierarchical bracketing encoding for dependency parsing as tagging, reducing label complexity and supporting non-projective trees more efficiently, achieving competitive accuracy.

Contribution

It proposes an optimal hierarchical bracketing encoding that minimizes label usage and extends support to non-projective trees, improving over previous methods.

Findings

Uses only 12 labels for projective trees, fewer than previous 16-label encoding.

Supports arbitrary non-projective trees more compactly than prior encodings.

Achieves competitive accuracy across diverse treebanks.

Abstract

We present a family of encodings for sequence labeling dependency parsing, based on the concept of hierarchical bracketing. We prove that the existing 4-bit projective encoding belongs to this family, but it is suboptimal in the number of labels used to encode a tree. We derive an optimal hierarchical bracketing, which minimizes the number of symbols used and encodes projective trees using only 12 distinct labels (vs. 16 for the 4-bit encoding). We also extend optimal hierarchical bracketing to support arbitrary non-projectivity in a more compact way than previous encodings. Our new encodings yield competitive accuracy on a diverse set of treebanks.