The Interplay of Variant, Size, and Task Type in Arabic Pre-trained Language Models
This work addresses the challenge of selecting effective pre-trained models for Arabic NLP tasks, but it is incremental as it builds on existing methods for language variants and fine-tuning.
The study investigated how language variants, data sizes, and task types affect Arabic pre-trained language models, finding that variant proximity between pre-training and fine-tuning data is more critical than data size, with results used to define an optimized system selection model.
In this paper, we explore the effects of language variants, data sizes, and fine-tuning task types in Arabic pre-trained language models. To do so, we build three pre-trained language models across three variants of Arabic: Modern Standard Arabic (MSA), dialectal Arabic, and classical Arabic, in addition to a fourth language model which is pre-trained on a mix of the three. We also examine the importance of pre-training data size by building additional models that are pre-trained on a scaled-down set of the MSA variant. We compare our different models to each other, as well as to eight publicly available models by fine-tuning them on five NLP tasks spanning 12 datasets. Our results suggest that the variant proximity of pre-training data to fine-tuning data is more important than the pre-training data size. We exploit this insight in defining an optimized system selection model for the studied tasks.