Arabic extremism detection based on using different machine learning techniques

What is it about?

Extremist groups frequently use social media without much help from network providers to spread their message and attract new members. In order to ensure that extremist information can be quickly removed and its producers are prevented from accessing networks, new methods are being developed to identify it. the aim of detecting extremism in Arabic texts on social media is to identify and stop hate speech and radical ideologies and violent content that threatens social cohesiveness and national security by inciting violence. Our goal in this project was to gather an Arabic-language tweet dataset and use machine learning (SVM, Random Forest) and Deep learning (Bi-LSTM) to identify extremist content. Bi-LSTM surpassed other traditional machine learning algorithms in terms of accuracy reached to 95.88%.

Featured Image

Why is it important?

The importance of this paper based on detecting extremism

Perspectives