A Watchdog Proposal to a Personal e-Health Approach

Abstract

During everyday activities, stress and anxiety are considered factors that influence users’ behavior and, if recurrent, can bring different risks to the individual’s health. This work proposes an architecture to assist the users in detecting and controlling emotions during their daily activities. The proposal was implemented using smartbands as body sensors to collect the data, machine learning algorithms to detect moments of stress, and a smartphone app for monitoring the user’s environment. The proposal evaluation was carried out by collecting real data from a user for one year. Based on that, we detected frequency peaks and, using the location information, enriched the data to send recommendations. The results indicates the feasibility of the proposal since it was possible to identify stressful moments considering the environment that the user wanted to be monitored.