Representing scents: An evaluation framework of scent-related experiences through associations between grounded and psychophysiological data

Abstract

This study introduces an empirical approach for assessing human scent-related experiences within the field of Human-Computer Interaction (HCI). We labeled 43 fragrances based on grounded collective experience, incorporating semantic and impression-based data. Furthermore, we collected comprehensive psychophysiological data, including electroencephalogram (EEG), electrobulbogram (EBG), electrocardiogram (ECG), and facial dynamics captured by a camera, from participants who experienced the scents. By computing scent-wise similarity and correlating both grounded and psychophysiological scent spaces, we identified associations between them, demonstrating the potential of this approach to enhance our understanding of scent-related experiences. Additionally, we propose an iterative evaluation framework to refine the design of smell-based interactions and we conduct a real-life study to validate this framework.