From FusHa to Folk: Exploring Cross-Lingual Transfer in Arabic Language Models
This work addresses limitations in Arabic NLP for dialectal applications, but it is incremental as it analyzes existing transfer issues without proposing new solutions.
The study investigated cross-lingual transfer in Arabic language models from Modern Standard Arabic to dialects, finding that transfer is possible but varies disproportionately across dialects, partially explained by geographic proximity, with evidence of negative interference in models trained on all dialects.
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.