This repo represents model developed for Irony and sentiment detection in Arabic tweets in WANLP shared tasks on sarcasm and sentiment detection in Arabic tweets

One of the problems faced concerning Arabic fake news detection is the scarcity of Arabic datasets. We believe it is important to available a dataset is written in the Arabic language explicitly created for this domain. The existing studies of Arabic fake news detection are limited because most of the dataset has not been available to access for…

Quick and Simple Approach for Detecting Hate Speech in Arabic Tweets.

an Arabic Twitter Corpus for Crisis Events

We utilized a pre-trained model to classify Arabic text. After conducting extensive research, we found that MarBERT was the best model for classifying Arabic offensive tweets. It focuses on dialectal Arabic (DA) and Modern Standard Arabic (MSA). The competition involves two shared sub-tasks: detecting whether a tweet is offensive or not; and det…

Sentiment Analysis of Arabic and Persian Tweets Following the Assassination of Qasem Soleimani