Virtual Agent Emotion Modeling for Interactive Human Training Bots

Computer Generated Solutions (CGS) is developing a chatbot to train health-related customer support human
agents. In this research, an intelligent sentiment-based chat bot is to be modeled and designed to act as a client
to train the client-facing representatives who would be able to help clients more effectively. The main objectives
of this internship are to (1) conduct literature review on virtual agents emotion modeling (2) design a novel
framework and emotion model using fuzzy systems, (3) to design and develop a trainer bot using language
understanding and emotional modeling, (4) to integrate the solution with the current CGS virtual agent avatar and
(5) to document the process and develop manuals for non technical experts in CGS to design intelligent bots.
Research outcomes of these studies provides the community with a systemic comparison of different ways that a
virtual agent could perceive and express emotions to act more naturally in a training setting.

Faculty Supervisor:

Alireza Manashty

Student:

Ghazal Erfani

Partner:

Computer Generated Solutions Canada Ltd.

Discipline:

Computer science

Sector:

Professional, scientific and technical services

University:

University of Regina