Many Languages, One Parser

TL;DR

This paper presents a unified multilingual dependency parser that leverages language-specific features and universal linguistic properties, enabling effective parsing across multiple languages with limited annotated data.

Contribution

The authors introduce a single multilingual parser that uses language-specific features and universal representations, improving cross-lingual parsing performance especially with limited data.

Findings

Performs well across languages with varying data sizes

Generalizes effectively based on linguistic universals

Outperforms strong baselines in multiple scenarios

Abstract

We train one multilingual model for dependency parsing and use it to parse sentences in several languages. The parsing model uses (i) multilingual word clusters and embeddings; (ii) token-level language information; and (iii) language-specific features (fine-grained POS tags). This input representation enables the parser not only to parse effectively in multiple languages, but also to generalize across languages based on linguistic universals and typological similarities, making it more effective to learn from limited annotations. Our parser's performance compares favorably to strong baselines in a range of data scenarios, including when the target language has a large treebank, a small treebank, or no treebank for training.