Modelling human-agent interaction based on affective and cognitive computing

Resumen

Emotions are a big part of what makes us human, and is an inherent part of living. In these last years there has been a growing discussion and awareness regarding emotions and their role in mental health and the general well being of people, in big part thanks to different platforms being used to discuss those issues more openly. Thus, we are able to more clearly see and understand the importance of emotions, and also how many people have sought out help when some of those emotions have become so overwhelming they have interfered with their day to day life. Many of those discussion surround the more traditional ways to address these issues, such as with different types of psychological therapy, or the use of medication prescribed by a licensed professional. While these professionals are essential to guide the process of working through some of these emotional block, and to teach us how to better manage our emotions in order to avoid falling into darker corners of our mind, adoption of technology for such endeavors is of interest. In a society that is growing more and more accustomed to technology being part of their daily lives, it provides a wide variety of tools and options for a variety of purposes to make life a little bit easier, like the use of voice-activated assistants, or chatbots to help with certain tasks online. This Ph.D. thesis is meant to provide a better understanding of how different people interact with those assistants by means of modelling two different types of agents (virtual and physical), and to provide a tool for therapists to use in their sessions to help with matters surrounding emotional management. With this, it is our hope to provide to the array of tools another approach that serves everyone that aims to work their emotional management, and improve their mental health.