Emotion Based Chatbot Using Deep Learning

The use of chatbots has been evolving rapidly in numerous fields in recent years which includes Education, Marketing, Health Care, Entertainment, etc. Chatbots utilise the technology of Artificial Intelligence (AI) to mimic human conversation. It is like a virtual assistant that is able to assist in conversation, answering questions, getting directions in the area of knowledge they are designed under. Chatbot can reduce customer service cost and handle multiple users at a time which makes it become more popular. But yet to accomplish a task in detecting and identifying emotion from users the chatbot interacts with. To address this issue, this paper provides a design of a chatbot, in which the chatbot is able to identify the user’s emotional status through the text provided by the users through the communication and then provide suggestions according to the input given. The emotion detection model will be trained using Artificial Neural Networks (ANN) and supervised learning. This chatbot is proposed to be used by i-CATS University College to help in the Marketing department to automate the process of answering FAQs and inquiries. However, a full chatbot implementation that replies to human queries is suggested that this research paper should shows to indicate how effective the chatbot can be. It is also suggested that strong implementation of chatbot must be considered for the improvement of this research in the future.