Question answering system-based chatbot for health care

Abstract

Chatbot helps to provide automated as well as instant output at the absence of human intervention. It is more essential in an emerging domain like health care to manage the emergency condition without the presence of medical experts. In this research, we are motivated to develop a health-care chatbot system to recognize diseases from user-provided health conditions or symptoms. This research helps to overcome the above-mentioned challenges in partially. Primarily, these challenges are introduced due to the rapid development of information and communication technology. On the other hand, the chatbot industry is rapidly growing while promising to cut the costs. Also, less involvement of domain experts and lack of automated information extraction system introduced more difficulties in this task. Hence, we have employed an unsupervised machine learning technique to build this chatbot. Additionally, we have prepared an experimental dataset that assists in validating the output of the proposed system. Primarily, this system recognized a set of diseases from the user given a set of symptoms and vice versa.