Human-AI Interaction Design Lab
Ting Liao, Ph.D.
Assistant Professor, School of Systems and Enterprises, Stevens Institute of Technology
Keywords: engineering design, human-centered design, human-computer interaction, socio-technical system, human-AI collaboration, trust in AI
Design of empathetic chatbot for better trust and engagement
We design virtual agents that deliver empathetic responses within the context of academic advising and investigate how agent behavior and user emotional states influence user engagement and trust. The findings enable more trustworthy and empathetic agents that potentially increase user retention and accordingly, make education and counseling more accessible.
A real-time framework to provide emotion-congruent responses in human-AI interactions
We design and develop a framework that utilizes both sentiment analysis and facial recognition to track user emotional states in real-time in the interaction between users and a virtual agent. The framework enables the agent to provide customizable and emotion-congruent responses when emotion acknowledgment is important, for example, in mental counseling.
Wearable data integrity detection and causation for effective human interaction
We investigate how human preference and security behavior impact data integrity and cyber risk for telehealth when using wearable devices under nextG networks. The proposed framework strengthens other existing methods for assessing cyber risks by accounting for human behavior. The output of this research will allow security engineers to monitor the telehealth system in real-time and protect user confidentiality and data integrity more efficiently.