Analysis of the Effects of Lifelike Characters on Computer-mediated Communication

The goal of this thesis is to clarify the effects of lifelike characters on computer mediated communications. In this research, lifelike characters are defined as personified agents and avatars which are represented as human-like characters with a face and facial expression on the computer screen. Most research within the field of personified agents has been focused on creating believable characters that act autonomously as we humans do, while the research focus on avatars has been creating systems and templates that correspond to a user’s intended emotions of gestures. However, the implementations of agents or avatars in the above research are based on two assumptions. The former assumption is that lifelike characters on the interface have positive effects on human-character interactions or human-human interactions via characters. There are arguments that personification of agents may cause users’ higher expectation in their perceived intellectuality than they actually have. The latter assumption is that lifelike characters and their facial expressions are recognized universally across cultures. There are only few studies that compare users’ interpretations of character’s facial expressions between two countries, and none among multiple countries. There should be research that verifies the effects of using lifelike characters on the interface, and whether the appearance of characters is equally interpreted across any cultures. With respect to the above research issues, this thesis verified the effects of using lifelike avatars in the following communication domains.