Subjective Importance and Computational Models of Emotions

In this paper, I shall attempt to do three things. First, I shall try to show that the standard characterization of subjective importance (which is widely recognized as crucial to the elicitation of emotions) is not powerful enough to fulfill the role that is usually attributed to it in accounts of emotions. Second, I shall discuss what the role of subjective importance in emotions is and how it might be computed. Finally, I shall suggest how, through the use of more specific constructs, this role can be realized in a cognitive account of emotion elicitation and intensification to make a system of computationally tractable rules a practical possibility.