Analysis of Perceptions towards Strong Emotions in Intercultural Communication

Abstract

In the field of human-robot interaction, we always pursue achieving a warmer robot. Obviously, emotion perceptions in a single cultural cannot meet needs when cultural blends become more common and widespread. This study focuses on cross-cultural groups, collecting voice responses with three languages (Chinese, English and Japanese) in two attitudes (strongly positive and strongly negative). We found out how exposures of intense feelings are affected by different linguistic contexts by clustering the data. The subjects could be roughly classified into several types assisted by comparing the degree of emotional exposure in their mother tongue and in non-native languages. Besides, we also found that the mastery of the other language is also an important factor in the degree of emotional exposure. We therefore suggest that when planning spoken human-robot interaction, people with multicultural background should be taken into consideration for a more comprehensive robot model.