Using Deep Learning Transformers for Detection of Hedonic Emotional States by Analyzing Eudaimonic Behavior of Online Users 

This project focuses on detecting hedonic emotional states by analyzing short text descriptions that reflect moments of happiness. The approach involves using deep learning transformer models to understand and classify these emotions. The system is designed to learn patterns from the HappyDB dataset, which contains thousands of happy moments written in natural language. Various deep learning algorithms such as LSTM, Bi-LSTM, DISTILBERT + LSTM AND ROBERTA AND BERT and XLNET are used to capture both sequential and contextual features of text data. A web application is developed using the Flask framework to allow users to register, log in, and enter their text for emotion detection. The interface processes the input and displays the predicted emotional state. This method demonstrates how advanced language models can be applied to detect emotions from brief textual inputs. The overall aim is to develop a reliable and interactive system that can understand user emotions and support further studies in affective computing and emotion-aware systems.

Keywords: Hedonic emotions, Deep learning, Transformers, LSTM, Bi-LSTM , BERT AND ROBERTA, XLNET, Flask, Text classification.