Identifying Cyber Extremism Sentiments using ROBERTA

It is critical to identify and mitigate cyber extremist sentiments in order to prevent terrorist acts and maintain national security. The use of online platforms to promote extremist ideologies and incite violence is referred to as cyber extremism. To detect and classify extremist content on the internet, machine learning models have been used, making the process more efficient and scalable. To identify cyber extremism sentiments using machine learning models, the model is typically trained on a large dataset of extremist and non-extremist materials. The model then learns to recognize patterns and features associated with extremism, such as the use of specific words and phrases, as well as image and video-based features like facial expressions and gestures. We use ROBERTA for this purpose in this study and compare its results to those of BERT and ALBERT. The use of machine learning to detect cyber extremism sentiments can improve law enforcement and other organizations' ability to detect and respond to potential threats, ultimately helping to keep communities safe.