From FusHa to Folk: Exploring Cross-Lingual Transfer in Arabic Language Models

TL;DR

This paper investigates how well Arabic language models trained on Modern Standard Arabic transfer to various dialects, revealing uneven transfer success influenced by geographic proximity and potential negative interference.

Contribution

It provides the first comprehensive analysis of cross-dialect transfer in Arabic models, highlighting limitations and challenges in dialectal diversity.

Findings

Transfer is possible but varies across dialects.

Geographic proximity influences transfer effectiveness.

Negative interference occurs in models supporting multiple dialects.

Abstract

Arabic Language Models (LMs) are pretrained predominately on Modern Standard Arabic (MSA) and are expected to transfer to its dialects. While MSA as the standard written variety is commonly used in formal settings, people speak and write online in various dialects that are spread across the Arab region. This poses limitations for Arabic LMs, since its dialects vary in their similarity to MSA. In this work we study cross-lingual transfer of Arabic models using probing on 3 Natural Language Processing (NLP) Tasks, and representational similarity. Our results indicate that transfer is possible but disproportionate across dialects, which we find to be partially explained by their geographic proximity. Furthermore, we find evidence for negative interference in models trained to support all Arabic dialects. This questions their degree of similarity, and raises concerns for cross-lingual transfer in Arabic models.