Especially in the current crisis of the COVID-19 pandemic and the lockdown it entailed, technology became crucial. Machines need to be able to interpret and represent human behavior, to improve human interaction with technology. This holds for all domains but even more so for the domain of student behavior in relation to education and psychological well-being.
This work presents the theoretical framework of a psychologically driven computing ontology, CCOnto, describing situation-based human behavior in relation to psychological states and traits. In this manuscript, we use and apply CCOnto as a theoretical and formal description system to categorize psychological factors that influence student behavior during the COVID-19 situation. By doing so, we show the added value of ontologies, i.e., their ability to automatically organize information from unstructured human data by identifying and categorizing relevant psychological concepts.
The already existing CCOnto was modified to automatically categorize university students’ state and trait markers related to different aspects of student behavior, including learning, worrying, health, and socially based on psychological theorizing and psychological data conceptualization.
The paper discusses the potential advantages of using ontologies for describing and modeling psychological research questions. The handling of dataset completion, unification, and its explanation by means of Artificial Intelligence and Machine Learning models is also discussed.