Enhancing tutoring systems by leveraging tailored promptings and domain knowledge with Large Language Models
Project Overview
The document explores the transformative role of generative AI, especially through Intelligent Tutoring Systems (ITS), in enhancing Computer-Based Learning (CBL). It emphasizes the utilization of Large Language Models (LLMs) to create personalized learning experiences by adapting to individual student needs and offering customized feedback. Key findings indicate that integrating skill-aligned feedback with Retrieval Augmented Generation (RAG) significantly improves the effectiveness of these tutoring systems. However, despite these advancements, challenges persist in addressing diverse learning styles and ensuring the delivery of real-time feedback to students. Overall, the document underscores the potential of generative AI to revolutionize educational practices while acknowledging the obstacles that need to be overcome to fully realize its benefits.
Key Applications
Intelligent Tutoring Systems (ITS) using Large Language Models (LLMs)
Context: Educational context focusing on computer science programming for students with varying skill levels.
Implementation: Developed a system that profiles student skills and integrates feedback into LLMs using structured prompt engineering.
Outcomes: Improved learning experiences through personalized, context-aware feedback leading to better engagement and understanding.
Challenges: Difficulty in accommodating diverse learning styles and providing real-time feedback.
Implementation Barriers
Technical Barrier
Challenges in delivering real-time, context-aware feedback.
Proposed Solutions: Integrating skill-aligned feedback via Retrieval Augmented Generation (RAG) and improving prompt engineering.
Pedagogical Barrier
Inability to customize ITS to accommodate diverse learning styles.
Proposed Solutions: Profiling student skills and continuously integrating feedback based on student performance.
Project Team
Mohsen Balavar
Researcher
Wenli Yang
Researcher
David Herbert
Researcher
Soonja Yeom
Researcher
Contact Information
For information about the paper, please contact the authors.
Authors: Mohsen Balavar, Wenli Yang, David Herbert, Soonja Yeom
Source Publication: View Original PaperLink opens in a new window
Project Contact: Dr. Jianhua Yang
LLM Model Version: gpt-4o-mini-2024-07-18
Analysis Provider: Openai