Transition-based Bubble Parsing: Improvements on Coordination Structure Prediction

TL;DR

This paper introduces a transition-based bubble parser that improves coordination structure prediction by explicitly modeling coordination boundaries and internal relationships, achieving state-of-the-art results on English datasets.

Contribution

It presents a novel transition system and neural models for parsing bubble-enhanced structures, advancing dependency-based syntactic analysis.

Findings

Outperforms previous methods on English Penn Treebank and GENIA corpus

Particularly effective on complex coordination structures

Achieves state-of-the-art accuracy in coordination prediction

Abstract

We propose a transition-based bubble parser to perform coordination structure identification and dependency-based syntactic analysis simultaneously. Bubble representations were proposed in the formal linguistics literature decades ago; they enhance dependency trees by encoding coordination boundaries and internal relationships within coordination structures explicitly. In this paper, we introduce a transition system and neural models for parsing these bubble-enhanced structures. Experimental results on the English Penn Treebank and the English GENIA corpus show that our parsers beat previous state-of-the-art approaches on the task of coordination structure prediction, especially for the subset of sentences with complex coordination structures.