Pre-Training BERT on Arabic Tweets: Practical Considerations
This work addresses the problem of improving NLP performance for Arabic dialects and social media, though it is incremental as it builds on existing BERT methods.
The researchers tackled the challenge of pretraining BERT models for Arabic NLP by developing five models with variations in training data size, mix of formal and informal Arabic, and linguistic preprocessing, achieving new state-of-the-art results on several downstream tasks.
Pretraining Bidirectional Encoder Representations from Transformers (BERT) for downstream NLP tasks is a non-trival task. We pretrained 5 BERT models that differ in the size of their training sets, mixture of formal and informal Arabic, and linguistic preprocessing. All are intended to support Arabic dialects and social media. The experiments highlight the centrality of data diversity and the efficacy of linguistically aware segmentation. They also highlight that more data or more training step do not necessitate better models. Our new models achieve new state-of-the-art results on several downstream tasks. The resulting models are released to the community under the name QARiB.