The AI-based Classroom Teaching Simulator (ACTS) is a conversational agent designed to help teachers practice research-based instructional strategies in content-based scenarios. The ACTS web interface has both chat functionality as well as a dynamic representation that teachers can use to discuss a task with a virtual student. ACTS relies on recent advances in natural language processing, deep learning, and uncertainty quantification and features human-in-the-loop functionality. 

This project brings together researchers from the School of Education and Human Development and the School of Data Science at the University of Virginia, as well as scholars from James Madison University with the support of Desmos PBC. The project is supported generously by the National Science Foundation, the Robertson Family Foundation, the Learning Tools Competition, and 4-VA: Advancing the Commonwealth grants.