Learning user well-being and comfort through smart devices

Abstract

The growth of concepts such as Intelligent Environments and Internet of things allows us to understand the habits of users and consequently act to improve people’s daily lives. Through information gathering, it is thus possible to gather patterns about different kinds of human behavior and consequently build a learning model with predictive capabilities. In addition, there are increasing concerns from large companies about the influence, positive or negative, that aspects such as comfort and well-being have on the behavior and health of the population. In fact, as human beings, we are greatly influenced by the environment in which we are inserted. There are therefore conditions in a place that give us certain levels of comfort that will eventually interfere with our well-being. However, it is difficult to identify which of these factors are relevant and how they intervene in our daily lives. Also, the habits we adopt as a result of the routines we follow can contribute to improving or worsening any of these indicators With the help of the various types of sensors present, for example, in the smart devices (smartphones, smartwatches, wristbands), it is increasingly possible to collect information on these factors, easily and comprehensively. In this sense, firstly the main objective of this dissertation is thus to collect data on factors that may influence the user in order to create a user profile. These factors can be inferred through its interests, the visited locations, and its main activities. This objective involves a large-scale analysis, where there are no geographical restrictions. Furthermore, the study will be independent of the type of space (open or closed) that is explored. In that way, the perspective that will be used is from the user. Then there is an exploration of the data so that some intelligence can be inferred, and in this sense, build a mobile application capable of providing smart notifications based on user needs.