Analyzing the Effect of Linguistic Similarity on Cross-Lingual Transfer: Tasks and Experimental Setups Matter

TL;DR

This study investigates how linguistic similarity influences cross-lingual transfer performance across 263 diverse languages and three NLP tasks, revealing that the impact varies based on task, representations, and similarity definitions.

Contribution

It extends previous research by analyzing a large, diverse set of languages and multiple NLP tasks, providing a comprehensive understanding of factors affecting transfer success.

Findings

Linguistic similarity's effect varies with NLP task.

Input representations influence transfer outcomes.

Definition of linguistic similarity impacts transfer performance.

Abstract

Cross-lingual transfer is a popular approach to increase the amount of training data for NLP tasks in a low-resource context. However, the best strategy to decide which cross-lingual data to include is unclear. Prior research often focuses on a small set of languages from a few language families and/or a single task. It is still an open question how these findings extend to a wider variety of languages and tasks. In this work, we analyze cross-lingual transfer for 263 languages from a wide variety of language families. Moreover, we include three popular NLP tasks: POS tagging, dependency parsing, and topic classification. Our findings indicate that the effect of linguistic similarity on transfer performance depends on a range of factors: the NLP task, the (mono- or multilingual) input representations, and the definition of linguistic similarity.