Mental states of users of cognitive interaction systems
This dissertation aims to develop the foundation for the automatic adaptation of cognitive interaction systems to individual, dynamically changing, mental user states, such as the degree of attention, workload or distraction. Based on continuously recorded bio-signals from eye movements, brain activities, speech and muscle activities, machine learning based methods are to be developed to interpret user states using contextual knowledge. It should also be investigated whether results from lab conditions can be transferred to measurements in the field using mobile devices.