USING EVERYDAY LIFE AUDIO SIGNALS TO GENERATE AUTOMATED MOOD ASSESSMENTS FOR ADAPTIVE SYSTEMS
Whereas decision making is a frequent everyday life task, its scientific investigation is mainly done in lab studies. However, progress in mobile technology in recent years enables studying such processes in daily life, using digital phenotyping, smartphone technology and wearables. This dissertation project will focus on affect, one of the most central determinants of adaptive systems. Leveraging recent developments from affective computing, we will extract sentiment and voice features of spoken language from daily life assessments as input for adaptive systems. We will partly focus on mental health populations to achieve maximum differences in experienced affect in everyday life.