The Impact of Social Media Addiction on The Attitude toward Extremism ‎Among University Students A Cross-Cultural Analytical Study Using Artificial Intelligence Techniques

     The current study aimed to investigate the impact of several demographic factors such as gender and academic specialization (scientific vs. literary) on the relationship between social media addiction and attitude toward extremism. It also aimed to examine the influence of cultural differences among Saudi, Egyptian, and Sudanese nationalities on both social media addiction and attitude toward extremism. Additionally, the study aimed to assess the effectiveness of using machine learning models in predicting attitude toward extremism based on social media addiction. Furthermore, it sought to identify which factors of social media addiction are the most predictive of attitude toward extremism. Finally, the study aimed to determine the suitability of various machine learning models in modeling and predicting the relationship between social media addiction and attitude toward extremism. The study sample consisted of (352) college students (123 males, 229 females) with an age mean of 20.1 years and a standard deviation of 0.99, selected by available method. The study tools included the social media addiction scale by Shahnawaz and Rehman (2020), translated and adapted into Arabic by the current researchers, and the attitude toward extremism scale by Ozer and Bertelsen (2018), also translated and adapted into Arabic by the current researchers. The study yielded several results, including the following: Demographic analyses highlighted differences in model performance based on gender and academic specialization. It is worth noting that the models achieved perfect classification accuracy for male participants and for both scientific and literary disciplines, indicating that attitude toward extremism may manifest differently across these demographic groups. Also, Classification results based on nationality emphasized the influence of cultural factors. It was also found that Random Forest analysis revealed the pivotal role of addiction factors in predicting attitude toward extremism among college students. notably, factors such as relapse and conflict resulting from social media addiction emerged as highly influential predictors, underscoring the importance of these aspects in understanding attitude toward extremism. Additionally, tolerance, salience, withdrawal, and mood modification factors were found to significantly contribute to predicting attitude toward extremism. These results confirm a multifaceted and intricate relationship between individual social media addiction factors and attitude toward extremism. Moreover, the analysis using SVM and ANN models revealed a clear picture of the predictive capabilities of social media addiction and demographic variables on attitude toward extremism. While both models showed reasonable accuracy, the ANN model demonstrated slightly better performance, highlighting its ability to capture complex patterns in the data.