Deep learning model for identification of customers satisfaction in business

Abstract

Communication is the key to expressing one’s feelings and ideas evidently. Recognition of the emotional state of a speaker is a significant step in making the human machine communication more natural and approachable. The knowledge behind creating this work was to make a deep learning model that might identify the emotions from their speech. This can be used by multiple industries to offer different services like marketing company signifying you to buy products based on your emotions, automotive industry can detect the customer’s emotions and adjust the speed of self-directed cars as required to avoid any collisions, etc. The proposed system has involved classifying the emotion into angry, calm, fear, happy and sad categories from the audio signals using classifier algorithms like MultiLayer Perceptron and Convolutional Neural Network. The customer satisfaction is very important to enhance the business. By using this model the customer can identify the customer satisfaction. The suggested methods open the door for a real-time prototype for customer speech emotion recognition with open-source features to boost business profitability.