Greg Nicholas
Undergraduate Research Assistant
Computer Science Dept. at the University of Pittsburgh

Under construction. (indefinitely)
As for the bare bones (in bare HTML):

I'm an undergraduate student at the University of Pittsburgh who has been fortunate enough to find a job as a research assistant under Dr. Diane Litman here at Pitt. This basically means that I get to pretend I'm a grad student in my spare time.

My research concerns improving how well computerized tutors sense emotions in their students. By analyzing features in student speech to look for emotional cues, the tutor might be able to adapt its lesson plan.

Why would this help things?

Imagine that a student answers a question correctly but has a very uncertain tone in his or her voice. If the tutor could sense this, it could continue to reinforce the idea to build more confidence in the student's understanding of the concept.

As a more practical example, if a tutor can sense a student's extreme anger and frustration with its software, it could perhaps act super nice to the student to avoid him or her throwing the monitor onto the floor. :-)

For more info, check out these links:

ITSPOKE main page - The research group I'm working with.
My SLT-2006 paper - My first published paper, "Exploiting Word-level Features for Emotion Prediction", by myself, Mihai Rotaru, and Diane Litman.