PNNL

Abstract

Attempts at interface agent personalization are usually aimed athelping the user perform a task or service. For example,scheduling of appointments, inspection of messages, discoveringitems of interest and different forms of negotiation. While this isvery noble undertaking, it makes assumptions about the level oftrust and credibility a user may place in such an agent in a realworld setting. If Microsoft’s experiments in social user interfacesteach us anything, it is that a ‘one size fits all’ solution does nottruly engage the user and encourage reuse. This ‘relationshipmanagement’ between the user and the character begins when thetwo first meet. As with human-human relationships, firstimpressions are essential. Instead of looking at the functionalaspects of the relationship, we believe the characters embodimentis the best place to start the personalization process. We describea study in which participants from several different age ranges,genders and ethnic groups were asked their preference ofanthropomorphic character, based on a cooperative computertask. We found that participants generally selected charactersfrom the same ethic group as themselves and that almost allparticipants selected a young character (instead of a middle agedor elderly character). No significant preference was found forcharacter gender.