TutorLLM: Customizing Learning Recommendations with Knowledge Tracing and Retrieval-Augmented Generation

Zhaoxing Li, Jindi Wang, Wen Gu, Vahid Yazdanpanah, Lei Shi, Alexandra I. Cristea, Sarah Kiden, Sebastian Stein

Abstract: The integration of AI into education enables more flexible and effective learning. Large Language Models (LLMs) such as ChatGPT offer broad topic coverage but lack personalization and may generate irrelevant or inaccurate content. To address these challenges, we propose TutorLLM, a personalized learning system that combines Knowledge Tracing (KT) and Retrieval-Augmented Generation (RAG). TutorLLM tailors responses based on each student’s learning state, predicted by the MLFBK KT model, and improves relevance using a Scraper component for context retrieval. Implemented as a Chrome plugin, TutorLLM was evaluated in a two-week field study with undergraduate students, demonstrating a 10% increase in user satisfaction and a 5% improvement in quiz scores compared to general LLMs.

External IDs:dblp:conf/interact/LiWGYSCKS25