The individual Adaption Module (iAM): A framework for individualization and calibration of companion technologies
The authors of the present study argue for the notion that assistant systems of the future will become “companion systems”, i.e., systems that adapt to individualspecificity of the user (i.e., to one's emotion patterns, personality, preferences, etc.). Due to the individual-specificity of emotion patterns, individualization and calibration processes will become necessary. In the present paper, an introduction to companion systems will be given, while placing special focus on the significance of automatic emotion recognition. Resulting problems for a transsituationally robust feature selection of signals and consequences on classification rates will be presented. In the end, a procedure for a potential structural individualization or calibration process for companion systems will be presented to find robust individual features. The vision is that robust individual-specific features and
multiple personality trait variables will be stored in an individual Adaptation Module (iAM). Such module will be outlined in the present article.