AraBERT
A language model developed by the AUBMIND Lab that pre-trains a BERT architecture on Arabic text. Training data covers approximately 24 GB of Modern Standard Arabic, including Arabic Wikipedia and news corpora. AraBERT outperforms multilingual BERT on Arabic morphological segmentation and semantic representation tasks.
Timeline (last 90 days)
النماذج المتخصصة مثل AraBERT وCAMeL تتعامل مع التعقيد الصرفي للعربية بشكل أفضل لأنها دُرِّبت على بيانات عربية أصلية وتضم طبقات تجزئة مُهيَّأة للبنية الاشتقاقية للعربية
The technical backbone of the framework is AraBERT — a Transformer model pre-trained for Arabic.
AraBERT is a language model pre-trained for Arabic by the AUBMIND Lab based on the BERT architecture.
AraBERT's training data sources include Arabic Wikipedia, news corpora, and other materials totaling approximately 24 GB of Modern Standard Arabic text.
Compared to multilingual BERT (mBERT), AraBERT has significant optimizations for Arabic morphological segmentation and semantic representation, especially in morphological restoration and root analysis.
All Facts (5)
النماذج المتخصصة مثل AraBERT وCAMeL تتعامل مع التعقيد الصرفي للعربية بشكل أفضل لأنها دُرِّبت على بيانات عربية أصلية وتضم طبقات تجزئة مُهيَّأة للبنية الاشتقاقية للعربية
50%UnverifiedThe technical backbone of the framework is AraBERT — a Transformer model pre-trained for Arabic.
50%UnverifiedAraBERT is a language model pre-trained for Arabic by the AUBMIND Lab based on the BERT architecture.
50%UnverifiedAraBERT's training data sources include Arabic Wikipedia, news corpora, and other materials totaling approximately 24 GB of Modern Standard Arabic text.
50%UnverifiedCompared to multilingual BERT (mBERT), AraBERT has significant optimizations for Arabic morphological segmentation and semantic representation, especially in morphological restoration and root analysis.
50%