doi: 10.18178/ijssh.2026.16.3.1290
Research on an Intelligent Tutoring System for Discrete Mathematics Integrating Knowledge Graph and LLM
Kun Luo
*
,
Yi Kuang
- School of Automation and Electronic Information, Xiangtan University, Xiangtan China
- Manuscript received April 11, 2026
- accepted June 11, 2026
- published August 31, 2026
Abstract
This study addresses the challenges of knowledge fragmentation and implicit computational thinking cultivation in discrete mathematics education. We propose an intelligent tutoring system that integrates knowledge graphs with large language models using a graph rag approach. The system adopts a three-tier architecture—data foundation, Knowledge Graph (KG) core, and intelligent empowerment—to bridge the gap between multi-source academic data and personalized tutoring. To explicitly support Computational Thinking (CT), the KG models the mapping between discrete mathematics concepts and CT dimensions such as abstraction and decomposition. Experimental results on domain-specific question answering tasks show that the KG-enhanced approach achieves an accuracy of 0.85 and an F1-Score of 0.87, significantly outperforming standard RAG and pure Large Language Model (LLM) methods. The findings demonstrate the technical feasibility of using structured knowledge to provide high-precision, reasoning-based support, laying a foundation for the systematic cultivation of computational thinking in intelligent education.
Keywords
- knowledge graph
- computational thinking
- Large Language Model (LLM)
- graph rag
How to Cite
Kun Luo and Yi Kuang, "Research on an Intelligent Tutoring System for Discrete Mathematics Integrating Knowledge Graph and LLM," International Journal of Social Science and Humanity, vol. 16, no. 3, pp. 124-129, 2026. https://doi.org/ 10.18178/ijssh.2026.16.3.1290
Copyright & License
Copyright © 2026 by the authors. This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited (CC BY 4.0).