Anxiety Detection During COVID-19 Using the Character Computing Ontology

Elaraby, N.; Bolock, A. E.; Herbert, C.; Abdennadher, S.

The lifestyle changes resulting from the COVID-19 pandemic increased the risk of suffering from anxiety and depression. We need the cooperation between psychologists and computer scientists to provide technology solutions to help mitigate negative mental well-being early on. CCOnto is an integrated ontology that models the interactions between behavior and character states and traits in specific situations following the framework of the inter-disciplinary domain of Character Computing. CCOnto is parts of an going research cooperation between computer scientists and psychologists for creating character-based interactive applications. The knowledge represented in the ontology is modular separating core knowledge from rules. In previous work, the rules were extracted from the literature. In this paper, we present an approach for generating rules from existing datasets. The main contribution of this paper is the generation of if/then rules from a dataset collected during the first lockdown. The rules are added to CCOnto in form of SWRL rules and used as a backend for an anxiety detection application. © 2021, Springer Nature Switzerland AG.