Using String Metrics to Improve the Design of Virtual Conversational Characters: Behavior Simulator Development...

Area of Science:

• Medical education
• Human-computer interaction
• Simulation technology

Background:

• Emergency waiting rooms are high-conflict environments requiring staff preparedness.
• The medical interview is a critical diagnostic and therapeutic tool, demanding complex communication skills.
• Virtual conversational characters (VCCs) are emerging as valuable tools in e-learning.

Purpose of the Study:

• To develop a modular architecture for a VCC-based behavior simulator.
• To create a tool for conflict avoidance training in hospital settings.
• To enable easy design and testing of training exercises.

Main Methods:

• Training exercises are designed as labeled, directed graphs.
• A system is developed to map user sentences to predefined sets for VCC responses.
• String metrics are compared to evaluate exercise design.

Main Results:

• A novel metric effectively identified ill-defined training exercises.
• The simulator's application in hospital environments was demonstrated.
• The metric provided insights into exercise clarity and effectiveness.

Conclusions:

• The VCC simulator, initially for emergency room staff, has broader applications.
• Potential uses in speech therapy are under exploration.
• The software facilitates the creation of effective conflict avoidance training scenarios.