Can polite computers produce better human performance

Abstract

We claim that the concept from human-human social interactions can be expanded and utilized to facilitate, inform, and predict human-computer interaction and perceptions. By expanding on a qualitative model of politeness proposed by Brown and Levinson we created a quantitative, computational model of etiquette that allows a machine to interpret and display politeness. The results from a human subject study show that the variables included in our model have important effects on subjects' decision making and performance in our experimental tasks. The results also demonstrate that variations in etiquette can result in objective, measurable consequences in human performance.