Psychological behavior Analysis in social media based on automated deep learning method

Suicide rates are notably higher in countries with higher levels of psychological stress. Depression and suicide may be avoided if signs of psychological stress are detected early. The widespread popularity and real-time social media data sharing in the last several decades have aided early diagnosis in big populations. The Deep Learning-based Automated Psychological Behavior Analysis Model (DLAPBAM) has been suggested to derive psychological stress from Twitter and Facebook usage. Focusing on deep learning and a transfer learning strategy that enhances an existing clinical text annotation collection, this study investigates methods for identifying suicide-related psychological stress. In this public opinion survey, these sources provide a wealth of information from which to draw. This study details a method for analyzing and synthesizing the sentiment of reviews that have been detected using deep learning methods, which may be quite useful when trying to anticipate the psychological behavior of a person. The experimental findings demonstrate an accuracy and prediction ratio improvement using the proposed DLAPBAM model over prior approaches. The experimental outcome has been accomplished, and the suggested methods prediction ratio (96.3 %), classification accuracy ratio (92.7%), efficiency ratio (96.8%), social media usage (97.9%), and error ratio (40.21 %).