On the Frailty of Universal POS Tags for Neural UD Parsers

TL;DR

This paper analyzes how the accuracy of universal POS tags affects neural dependency parser performance, revealing that high UPOS tagging accuracy is crucial and that gold tags significantly boost parsing results, with linguistic factors influencing tag quality.

Contribution

It provides a detailed analysis of UPOS tags' impact on neural parsers and identifies key linguistic aspects affecting tagging and parsing performance.

Findings

High UPOS accuracy is essential for optimal parsing.

Gold UPOS tags substantially improve parsing performance.

Linguistic features of tags influence parsing accuracy.

Abstract

We present an analysis on the effect UPOS accuracy has on parsing performance. Results suggest that leveraging UPOS tags as features for neural parsers requires a prohibitively high tagging accuracy and that the use of gold tags offers a non-linear increase in performance, suggesting some sort of exceptionality. We also investigate what aspects of predicted UPOS tags impact parsing accuracy the most, highlighting some potentially meaningful linguistic facets of the problem.