Typicality degrees to measure relevance of the physiological signals

Date
Journal Title
Journal ISSN
Volume Title
Publisher
SCITEPRESS Digital Library
Abstract
Physiological measures have a key advantage as they can provide an insight into human feelings that the subjects may not even be consciously aware of. However, modeling user affective states through pysiology still remains with critical questions especially on the relevant physiological measures for real-life emotionally intelligent applications. In this study, we propose the use of typicality degrees defined according to cognitive science and psychology principles to measure the relevance of the physiological features in characterizing user affective states. Thanks to the typicality degrees, we found consistent physiological characteristics for modeling user affective states.
Description
Paper presented at the International Conference on Physiological Computing Systems (PhyCS), Lisbon, Portugal.
Physiological measures have a key advantage as they can provide an insight into human feelings that the subjects may not even be consciously aware of. However, modeling user affective states through pysiology still remains with critical questions especially on the relevant physiological measures for real-life emotionally intelligent applications. In this study, we propose the use of typicality degrees defined according to cognitive science and psychology principles to measure the relevance of the physiological features in characterizing user affective states. Thanks to the typicality degrees, we found consistent physiological characteristics for modeling user affective states.
Keywords
Affective Computing, Physiological Signals, Machine Learning, Prototypes, Typicality Degrees, Gameplay
Citation