|
When humans collaborate face-to-face, they use a wide range of nonverbal signals that support the conversation process. Many of these signals are not available when people have conversations online by exchanging text messages, or are distorted when they attempt conversation through technology that attempts direct transmission, such as video conferencing. This lack of signal richness, introduces unwanted artifacts that lead to a disrupted conversational process. |
Hannes Vilhjalmsson |
|
This project presents a new approach to computer-mediated conversation that attempts to add the missing signals to the messages being passed, in a way that mimics a face-to-face encounter. A theoretical framework explains how a message can be automatically analyzed in terms of its communicative function, and then how behaviors, shown to support such functions in face-to-face conversations, can be added to an animated delivery of the message, featuring a automated graphical avatars. Spark is a new software architecture built on this framework that demonstrates the approach in an actual online collaboration system.
|
|