Virtual Human Personality Masks: A Human Computation Approach to Modeling Verbal Personalities in Virtual Humans

Modeling virtual humans that can exhibit realistic personalities is becoming increasingly important as virtual humans are being widely used for inter-personal skills education. We present Virtual Human Personality Masks, a system that combines human computation with the idea of using existing virtual humans to bootstrap the creation of other virtual humans to enable quick and easy generation of perceivable verbal personalities in virtual humans.

To evaluate this system, we created high and low verbosity-level variants of a virtual patient with symptoms of depression and conducted a user study with medical students split between two groups, each interacting with one of the two variants of the virtual patient. The participants’ perceived verbosity levels of the virtual patients indicated that not only did the virtual patients created using our system exhibit the intended personality in a perceivable manner, but also exhibited other related personality attributes in a manner that is consistent with the human personality theory analogs of verbosity.