ÌìÃÀmvÌìÃÀ

Modelling empathic behaviour in interactive systems for personalized interaction

Computational empathy is a recently emerging research area that builds on the premise of equipping intelligent virtual agents (IVAs) with the ability to act empathically—being able to recognize, resonate to and respond to the affective state of the users.

Being a sub-field of Affective and Social Computing, it shares the premise of pursuing the modeling of complex human behaviours as an attempt to both 1) enhance the interaction between humans and artificial agents, and 2) understand the complex emotion mechanisms in humans by testing theories and models in controlled environments that are difficult if not impossible using psychology or neuroscience experiments.

The long-term goal of Yalçın and her collaborators' research program is to advance the theory and practice of computational empathy research and build technologies that can adapt to the affective conditions of humans, by focusing on modeling capabilities that have not been formerly addressed.

Yalçın's research program will follow theory-driven modeling of the variety of empathic capacities (i.e., affect recognition, and theory of mind) as identified in my prior work, using AI models and techniques to implement these capacities in IVAs.

The proposed research will develop new theories, models, algorithms, and technologies that have a better understanding of the human condition during interaction, which will lead to effortless, natural, and more personalized interactions. This work will also advance the researchers' understanding of human behaviour through rigorous testing of models of empathic behaviour in a controlled and quantitative manner. Highly qualified personnel (HQP) will be trained in AI and HCI methods, particularly in natural interaction models.