A robot's experience of its user: Theory

Abstract

To make a robot that emotionally is more human-like, it should be capable of balancing simultaneously occurring tendencies of positive and negative affect. This paper is about the modeling of robot attitudes towards the user, using a fuzzy trade-off algorithm. Empirical research showed that in humans, involvement and distance are not mutually exclusive and that a range of factors contribute (e.g., ethics, affordances). A hypercube of fuzzy sets deals with combining these factors into an overall value for the involvement-distance trade-off, which subsequently can form the input for affective response selection (Bosse et al. 2008).