FACET: Framework for Agent-based Classroom Enhancement for Teacher

FACET is a teacher-centred LLM-based multi-agent system that supports teachers in developing personalized educational materials according to student characteristics, such as motivation, self-concept and performance.

problem

Teachers face increasingly diverse classrooms, yet despite the recognized importance of differentiation, limited time and resources make it difficult to translate personalization into everyday practice.

key features

• Multi-agent architecture: Incorporates multiple agents that simulate both students and the teacher
• Personalization: Delivers tailored learning materials based on individual students’ affective, motivational, and performance-related attributes
• Teacher-centred design: Supports teachers through AI-driven assistance while preserving their pedagogical autonomy and authority

outcome

Together with teachers, we are developing a teacher-centered, LLM-based multi-agent system designed to create AI-generated personalized teaching materials. Our approach goes beyond performance optimization by also taking into account students’ motivational and affective dimensions, ensuring a more holistic understanding of learning processes.

read more about the FACET framework

interestedt in co-creating the AI tool?

This research is part of our ongoing work in the ‘Humans and AI’ research area at Zuse Institute Berlin.