User Stress Level Detection From Social Networks Based On Social Interactions
Abstract
Psychological stress is the one of the main factors of the people’s health, it effect on working environment. Twitter is one of the best social media for sharing their regular feelings like emotions, angry, happy, daily activities, work status with their friends, employees and others. This platform needful to social media data for express the stress in social network. In this proposed method, detects the users stress state of social media users. The related data set developed by a hybrid model graph technique combined with dimensional reduction and Neural Network (CNN) based on user tweet content and social interaction. To find results by their friends in social media database by systematically study the user’s social interactions and correlation between user’s stress states. This method has given sophisticated results for improves the identification performance level by 8-15% in F1-score.