Fuzzy Logic Usage in Emotion Communication of Human Machine Interaction

Fuzzy Logic Usage in Emotion Communication of Human Machine Interaction

Zhe Xu (Bournemouth University, UK), David John (Bournemouth University, UK) and Anthony C. Boucouvalas (Bournemouth University, UK)
Copyright: © 2006 |Pages: 7
DOI: 10.4018/978-1-59140-562-7.ch036
OnDemand PDF Download:
$37.50

Abstract

As the popularity of the Internet has expanded, an increasing number of people spend time online. More than ever, individuals spend time online reading news, searching for new technologies, and chatting with others. Although the Internet was designed as a tool for computational calculations, it has now become a social environment with computer-mediated communication (CMC). Picard and Healey (1997) demonstrated the potential and importance of emotion in human-computer interaction, and Bates (1992) illustrated the roles that emotion plays in user interactions with synthetic agents. Is emotion communication important for human-computer interaction? Scott and Nass (2002) demonstrated that humans extrapolate their interpersonal interaction patterns onto computers. Humans talk to computers, are angry with them, and even make friends with them. In our previous research, we demonstrated that social norms applied in our daily life are still valid for human-computer interaction. Furthermore, we proved that providing emotion visualisation in the human-computer interface could significantly influence the perceived performances and feelings of humans. For example, in an online quiz environment, human participants answered questions and then a software agent judged the answers and presented either a positive (happy) or negative (sad) expression. Even if two participants performed identically and achieved the same number of correct answers, the perceived performance for the one in the positive-expression environment is significantly higher than the one in the negative-expression environment (Xu, 2005). Although human emotional processes are much more complex than in the above example and it is difficult to build a complete computational model, various models and applications have been developed and applied in human-agent interaction environments such as the OZ project (Bates, 1992), the Cathexis model (Velasquez, 1997), and Elliot’s (1992) affective reasoner. We are interested in investigating the influences of emotions not only for human-agent communication, but also for online human-human communications. The first question is, can we detect a human’s emotional state automatically and intelligently? Previous works have concluded that emotions can be detected in various ways—in speech, in facial expressions, and in text—for example, investigations that focus on the synthesis of facial expressions and acoustic expression including Kaiser and Wehrle (2000), Wehrle, Kaiser, Schmidt, and Scherer (2000), and Zentner and Scherer (1998). As text is still dominating online communications, we believe that emotion detection in textual messages is particularly important.

Complete Chapter List

Search this Book:
Reset