Feelflow: Delving into Emotional Depths with Generative AI

Abstract

Abstract: Significant progress has been made in sentiment analysis in recent decades, with a focus on textual data. Nonetheless, the scientific community has not done a great deal to investigate sentiment analysis in the context of audio. By applying a novel method of sentiment analysis to voice transcripts and concentrating on the subtle interpretation of emotions sent by distinct speakers during conversations, this study seeks to close this gap in knowledge. The main goal of this suggested research paper is to create an advanced sentiment analysis system that can communicate with several users in a seamless manner while identifying and assessing the emotional content that each user is conveying through their audio inputs. Advanced approaches like Recurrent Neural Networks (RNN), Long Short Term Memory(LSTM), Teacher Forcing , Encoder- Decoder Model, Tokenization, Gated Recurrent Units (GRU),(Bidirectional Encoder Representations from Transformers),Gradient Boosting Machine.