Classroom Simulacra: Building Contextual Student Generative Agents in Online Education for Learning Behavioral Simulation

2025-02-04

AI Summary

Existing student simulation methods overlook the dynamic regulatory role of course materials on student learning behaviors, hindered by the scarcity of fine-grained course data and insufficient modeling capacity for long educational texts. Method: We propose a Transferable Iterative Reflection (TIR) module—the first framework to jointly model classroom-level learning dynamics and inter-student behavioral correlations in both prompt engineering and fine-tuning paradigms for large language models (LLMs). Leveraging custom educational systems to collect fine-grained behavioral logs, TIR integrates course material embeddings, transfer learning, and interactive virtual student agents. Contribution/Results: Experiments demonstrate that TIR-driven LLMs significantly outperform classical deep learning models under limited demonstration samples, accurately reproducing temporal evolution of learning performance and group-level behavioral correlations. This advances the practical deployment of “digital twins” in online education.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Cognitive Modeling & Cognitive Systems: Simulating Human Behavior
• Multiagent Systems: Agent-Based Simulation and Emergent Behavior

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Search and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Student simulation supports educators to improve teaching by interacting with virtual students. However, most existing approaches ignore the modulation effects of course materials because of two challenges: the lack of datasets with granularly annotated course materials, and the limitation of existing simulation models in processing extremely long textual data. To solve the challenges, we first run a 6-week education workshop from N = 60 students to collect fine-grained data using a custom built online education system, which logs students' learning behaviors as they interact with lecture materials over time. Second, we propose a transferable iterative reflection (TIR) module that augments both prompting-based and finetuning-based large language models (LLMs) for simulating learning behaviors. Our comprehensive experiments show that TIR enables the LLMs to perform more accurate student simulation than classical deep learning models, even with limited demonstration data. Our TIR approach better captures the granular dynamism of learning performance and inter-student correlations in classrooms, paving the way towards a ''digital twin'' for online education.

Problem

• Simulate student learning behaviors accurately
• Overcome limitations in processing long textual data
• Capture granular dynamism in online education

Innovation

• Custom online education system
• Transferable Iterative Reflection module
• Enhanced large language models