Prototype of a virtual assistant for stress diagnosis using Artificial Intelligence

Abstract

This project aims to develop an intelligent virtual assistant of the Chatbot type to detect the level of stress, which works as a conversation based on questions and answers, that is, a bot-human-bot interaction. To this end, a descriptive and applied research (R+D) was carried out. The hypothetical- deductive scientific method was use based on the assumption that the level of a person’s stress symptoms is detect with the help of a virtual assistant. First, a study was carried out to know the symptoms of the disease, through bibliographic review and interview with specialists in psychology, and then a technological development was carried out using the SCRUM methodology considering three events: Sprint Planning, Daily Scrum and Sprint Review. The development tool used was Watson Assistant, in which created a prototype of a Chatbot capable of detecting whether a person suffers from stress or not, using the instrument published by the Headington Institute. A pilot test was carried out with students from the Faculty of Computational System of the Technological University of Panama and Faculty of medicine of the University of Panama that helped train the Chatbot, and with the help of the psychology specialists, it was validated if the answers given by the Chabot are correct.