Socio-Emotional Computational Analysis of Propaganda Campaigns on Social Media Users in the Middle East

Society has been signifcantly impacted by social media platforms in almost every aspect of their life. This impact has been efectively formulating people's global mindsets and opinions on political, economic, and social events. Such waves of opinion formation are referred to as propagandas and misinformation. Online propaganda infuences the emotional and psychological orientation of people. The remarkable leaps in Machine Learning models and Natural Language Processing have helped in analyzing the emotional and psychological efects of cyber social threats such as propaganda campaigns on diferent nations, specifcally in the Middle East, where rates of disputes have risen after the Arab Spring and the ongoing crises. In this paper, we present an approach to detect propagandas and the associated emotional and psychological aspects from social media news headlines that contain such a contextualized cyber social attack. We created a new dataset of headlines containing propaganda tweets and another dataset of potential emotions that the audience might endure when being exposed to such propaganda headlines. We believe that this is the frst research to address the detection of emotional reactions linked to propaganda types on social media in the Middle East.