Improvement in Multi-User Communication Systems Using an Avatar's Facial Expression Features

Abstract

We investigated how an avatar's facial expression function affects users' motivation to engage in conversation and ability to understand a partner's emotion during online chat communication and attempted to improve the chat system based on the usability ratings of this function. In Experiment 1, 15 pairs of university students chatted in English using both an emoticon-input facial expression system and a non-facial expression system. The evaluation of the systems' usability indicated that the emoticon-input system was not superior to the non-facial expression system. The emoticon-input system was thus modified to an icon-input system to improve its usability. In Experiment 2, 16 pairs of students chatted in Japanese using both the icon-input facial expression system and the non-facial expression system and guessed their partner's emotional states during the chat. Participants found it easier to guess their partner's emotions when using the facial expression system. We discuss the outcomes and limitations of our experiments for future research.