Implicit affective profiling of subjects based on physiological data coupling

Abstract

Evidence from neuroscience and psychology supports the idea that synchronization among various sub-components of the organism is the key-point to emotional processes. This study investigates the way signals from the central nervous system interact with signals from the peripheral nervous system during emotional processes elicited by music clips. In particular, the study reveals that coupling between skin conductance response (SCR) and electroencephalography (EEG) of the temporal lobe is significantly stronger compared to the coupling between SCR and EEG captured from other parts of the cortex (p < 0.01) during music clips. It is also shown that the coupling between SCR and EEG of the temporal lobe increases significantly during high arousing music clips with respect to low arousing ones ( p < 0.01), as well as during highly pleasant ( p < 0.01) or very unpleasant ( p < 0.01) music clips with respect to neutral ones. The potential of including such coupling between EEG and SCR in an affective brain-computer interface (BCI) is discussed. Moreover, the study presents a clustering scheme to profile human subjects, based on the coupling between SCR and EEG of the temporal lobe, shown to be related to their underlying emotional characteristics.

Keywords:

• EEG
• EDA
• phase-amplitude coupling
• affective computing
• affective profiling

Acknowledgements

The research leading to these results has been performed in the framework of the Swiss National Foundation for Scientific Research [FN 200,020-132,673-1], [FN 200,020-149,259-1].

Notes

1. http://www.cis.hut.fi/projects/ica/fastica/.