An Innovative Emotion Recognition and Solution Recommendation Chatbot
Ashlin Deepa R N, Prathyusha Karlapati, Mrunhaalhini Reddy Mulagondla, Pavitra Amaranayani, Anika Pranavi Toram
Abstract
The proposed chatbot for emotion recognition and solution recommendation system is a web-based application that aims at helping people to handle their emotions without any external assistance. In today's world, pressure and stress on the professional front, insecure relationships, and other factors produce a lot of mental turmoil, which many prefer not to discuss with others. The proposed system simulates a one-on-one interaction of the user with the chatbot through images, category selection, and text data describing the mood of the person. The user's text description of emotion is analyzed using a variety of machine learning algorithms and parameters, with Random Forest proving to be the most precise in recognizing emotion with accuracy and F1 score of 97.55 and 0.969, respectively. This facilitates recognizing subtle and hidden emotions to recommend better ways of handling the emotions. The proposed chatbot uses state of art technology to analyze the mood of the user using multiple inputs and recommends different ways of controlling the emotion.