AraBERT: Transformer-based Model for Arabic Language Understanding

TL;DR

This paper introduces AraBERT, a transformer-based language model specifically pre-trained for Arabic, achieving state-of-the-art results on various Arabic NLP tasks and providing resources to advance Arabic language understanding.

Contribution

The paper presents the first Arabic-specific BERT model pre-trained on a large corpus, outperforming multilingual BERT on multiple NLP tasks for Arabic.

Findings

AraBERT outperforms multilingual BERT on Arabic NLP tasks.

Pretrained AraBERT models are publicly available for research.

Achieved state-of-the-art results on several Arabic NLP benchmarks.

Abstract

The Arabic language is a morphologically rich language with relatively few resources and a less explored syntax compared to English. Given these limitations, Arabic Natural Language Processing (NLP) tasks like Sentiment Analysis (SA), Named Entity Recognition (NER), and Question Answering (QA), have proven to be very challenging to tackle. Recently, with the surge of transformers based models, language-specific BERT based models have proven to be very efficient at language understanding, provided they are pre-trained on a very large corpus. Such models were able to set new standards and achieve state-of-the-art results for most NLP tasks. In this paper, we pre-trained BERT specifically for the Arabic language in the pursuit of achieving the same success that BERT did for the English language. The performance of AraBERT is compared to multilingual BERT from Google and other state-of-the-art approaches. The results showed that the newly developed AraBERT achieved state-of-the-art performance on most tested Arabic NLP tasks. The pretrained araBERT models are publicly available on https://github.com/aub-mind/arabert hoping to encourage research and applications for Arabic NLP.