Coordinate Constructions in English Enhanced Universal Dependencies: Analysis and Computational Modeling

TL;DR

This paper improves the representation of coordinate constructions in Enhanced Universal Dependencies for English by creating a manually verified dataset, identifying errors, and demonstrating that machine learning methods outperform heuristic rules in dependency propagation.

Contribution

It introduces a large-scale manually annotated dataset for enhanced UD coordinate constructions and compares rule-based and machine learning methods, showing the superiority of the latter.

Findings

Machine learning-based propagation outperforms heuristic rules.

Manual annotation reduces systematic errors in dependency graphs.

Neural graph parsers improve coordinate dependency predictions.

Abstract

In this paper, we address the representation of coordinate constructions in Enhanced Universal Dependencies (UD), where relevant dependency links are propagated from conjunction heads to other conjuncts. English treebanks for enhanced UD have been created from gold basic dependencies using a heuristic rule-based converter, which propagates only core arguments. With the aim of determining which set of links should be propagated from a semantic perspective, we create a large-scale dataset of manually edited syntax graphs. We identify several systematic errors in the original data, and propose to also propagate adjuncts. We observe high inter-annotator agreement for this semantic annotation task. Using our new manually verified dataset, we perform the first principled comparison of rule-based and (partially novel) machine-learning based methods for conjunction propagation for English. We show that learning propagation rules is more effective than hand-designing heuristic rules. When using automatic parses, our neural graph-parser based edge predictor outperforms the currently predominant pipelinesusing a basic-layer tree parser plus converters.