Emotion analysis plays a crucial part in understanding the feelings of human beings. People convey a variety of sentiments, behavior, and emotions via their speech, which can have significant impacts. In most of the previous work, nearly all projects have focused on analyzing the expression based on only positive, negative and neutral classification while our project analyzes the proposed system by categorizing the text into emotion classes like joy, sadness, anger, fear, love and surprise.
While the solution is an ordinary classification model in NLP, we plan to experiment with using a pretrained model like BERT and other approaches such as LSTMs to benchmark and draw conclusions from it, along with model explainability. In this project we have worked on the entire lifecycle of an NLP task - Preprocessing, EDA, Model Training and Benchmarking and Model Explainability using lime.