This paper proposes a way to address the quantitative description of decision strategies for expert decision-makers in order to take into account the effects of personality, emotion or mood on decision-making. Most of common taxonomies used in popular models to decribe user profiles and behaviors seem to be difficult to apply in empirical cases. Instead we propose to rely on computational models inspired by cognitive psychology. These models and the related methodology allow to extract meaningful data structures from the behaviors of decisionmakers. This data can be used to propose robust definitions of decision-styles. We then discuss the impact of this refined modeling on decision-support system functionalities.