An Optimized Hybrid Neural Network Model for Detecting Depression among Twitter Users

The proposed work is to extensively evaluate if a user is depressed or not using his Tweets on Twitter. With the omni presence of social media, this method should help in identifying the depression of users. We propose an Optimized Hybrid Neural Network model to evaluate the user tweets on Twitter to analyze if a user is depressed or not. Where Neural Network is trained using Tweets to predict the polarity of Tweets. The Neural Network is trained in such a way that at any point when presented with a Tweet the model outputs the polarity associated with the Tweet. Also, a user-friendly GUI is presented to the user that loads the trained neural network in no time and can be used to analyze the users’ state of depression. The aim of this research work is to provide an algorithm to evaluate users’ sentiment on Twitter in a way better than all other existing techniques.