An AI Education Framework Based on LLM-Knowledge Graph for Personalized Learning

Zihao Liang; Yiou Wang; Mingjie Zhao; Sen Feng; Yunfan Zhang; Yiqun Zhang

doi:10.1109/CIS69366.2025.11433876

An AI Education Framework Based on LLM-Knowledge Graph for Personalized Learning

Zihao Liang
, Yiou Wang
, Mingjie Zhao
, Sen Feng
, Yunfan Zhang
, Yiqun Zhang^*

^*Corresponding author for this work

Department of Computer Science

Research output: Chapter in book/report/conference proceeding › Conference proceeding › peer-review

Abstract

Educational applications of Large Language Models (LLMs) face two key challenges: balancing their generative adaptability with factual reliability, and ensuring content accuracy within personalized learning paths. This paper presents LKPS (Large Language Model-Knowledge Graph for Personalized Learning System), an adaptive learning system that addresses these issues via two core innovations: (1) the SafeRAG framework, which uses structured prompting, multimodal cross-verification, and post-process verification to minimize hallucinations while maintaining content reliability; and (2) a hybrid path planner that integrates collaborative filtering with real-time knowledge tracing for personalized pathway optimization. In simulated multimodal machine learning scenarios, LKPS significantly outperforms traditional methods in both personalization and learning success, offering a viable pathway toward secure and deeply personalized educational AI.

Original language	English
Title of host publication	2025 21st International Conference on Computational Intelligence and Security (CIS)
Publisher	IEEE
Number of pages	6
ISBN (Print)	9798331550486
DOIs	https://doi.org/10.1109/CIS69366.2025.11433876
Publication status	Published - 12 Dec 2025
Event	2025 21st International Conference on Computational Intelligence and Security, CIS 2025 - Nanning, China Duration: 12 Dec 2025 → 15 Dec 2025

Conference

Conference	2025 21st International Conference on Computational Intelligence and Security, CIS 2025
Country/Territory	China
City	Nanning
Period	12/12/25 → 15/12/25

User-Defined Keywords

Educational AI
LLMs
Hallucination
Retrieval Augmented Generation
Personalized Learning

Access to Document

10.1109/CIS69366.2025.11433876

Cite this

@inproceedings{db00fee7bf724da79ec0033fc796b909,

title = "An AI Education Framework Based on LLM-Knowledge Graph for Personalized Learning",

abstract = "Educational applications of Large Language Models (LLMs) face two key challenges: balancing their generative adaptability with factual reliability, and ensuring content accuracy within personalized learning paths. This paper presents LKPS (Large Language Model-Knowledge Graph for Personalized Learning System), an adaptive learning system that addresses these issues via two core innovations: (1) the SafeRAG framework, which uses structured prompting, multimodal cross-verification, and post-process verification to minimize hallucinations while maintaining content reliability; and (2) a hybrid path planner that integrates collaborative filtering with real-time knowledge tracing for personalized pathway optimization. In simulated multimodal machine learning scenarios, LKPS significantly outperforms traditional methods in both personalization and learning success, offering a viable pathway toward secure and deeply personalized educational AI.",

keywords = "Educational AI, LLMs, Hallucination, Retrieval Augmented Generation, Personalized Learning",

author = "Zihao Liang and Yiou Wang and Mingjie Zhao and Sen Feng and Yunfan Zhang and Yiqun Zhang",

note = "This work was supported by the 2023 Guangdong Higher Education Teaching Reform Project (Grant No. 637), titled {\textquoteright}Exploration on the Teaching Mode of Frontier-Oriented AI Course Content Reconstruction{\textquoteright}. Publisher Copyright: {\textcopyright} 2026 IEEE; 2025 21st International Conference on Computational Intelligence and Security, CIS 2025 ; Conference date: 12-12-2025 Through 15-12-2025",

year = "2025",

month = dec,

day = "12",

doi = "10.1109/CIS69366.2025.11433876",

language = "English",

isbn = "9798331550486",

booktitle = "2025 21st International Conference on Computational Intelligence and Security (CIS)",

publisher = "IEEE",

address = "United States",

}

Liang, Z, Wang, Y, Zhao, M, Feng, S, Zhang, Y & Zhang, Y 2025, An AI Education Framework Based on LLM-Knowledge Graph for Personalized Learning. in 2025 21st International Conference on Computational Intelligence and Security (CIS). IEEE, 2025 21st International Conference on Computational Intelligence and Security, CIS 2025, Nanning, China, 12/12/25. https://doi.org/10.1109/CIS69366.2025.11433876

TY - GEN

T1 - An AI Education Framework Based on LLM-Knowledge Graph for Personalized Learning

AU - Liang, Zihao

AU - Wang, Yiou

AU - Zhao, Mingjie

AU - Feng, Sen

AU - Zhang, Yunfan

AU - Zhang, Yiqun

N1 - This work was supported by the 2023 Guangdong Higher Education Teaching Reform Project (Grant No. 637), titled ’Exploration on the Teaching Mode of Frontier-Oriented AI Course Content Reconstruction’. Publisher Copyright: © 2026 IEEE

PY - 2025/12/12

Y1 - 2025/12/12

N2 - Educational applications of Large Language Models (LLMs) face two key challenges: balancing their generative adaptability with factual reliability, and ensuring content accuracy within personalized learning paths. This paper presents LKPS (Large Language Model-Knowledge Graph for Personalized Learning System), an adaptive learning system that addresses these issues via two core innovations: (1) the SafeRAG framework, which uses structured prompting, multimodal cross-verification, and post-process verification to minimize hallucinations while maintaining content reliability; and (2) a hybrid path planner that integrates collaborative filtering with real-time knowledge tracing for personalized pathway optimization. In simulated multimodal machine learning scenarios, LKPS significantly outperforms traditional methods in both personalization and learning success, offering a viable pathway toward secure and deeply personalized educational AI.

AB - Educational applications of Large Language Models (LLMs) face two key challenges: balancing their generative adaptability with factual reliability, and ensuring content accuracy within personalized learning paths. This paper presents LKPS (Large Language Model-Knowledge Graph for Personalized Learning System), an adaptive learning system that addresses these issues via two core innovations: (1) the SafeRAG framework, which uses structured prompting, multimodal cross-verification, and post-process verification to minimize hallucinations while maintaining content reliability; and (2) a hybrid path planner that integrates collaborative filtering with real-time knowledge tracing for personalized pathway optimization. In simulated multimodal machine learning scenarios, LKPS significantly outperforms traditional methods in both personalization and learning success, offering a viable pathway toward secure and deeply personalized educational AI.

KW - Educational AI

KW - LLMs

KW - Hallucination

KW - Retrieval Augmented Generation

KW - Personalized Learning

UR - https://ieeexplore.ieee.org/document/11433876/

U2 - 10.1109/CIS69366.2025.11433876

DO - 10.1109/CIS69366.2025.11433876

M3 - Conference proceeding

SN - 9798331550486

BT - 2025 21st International Conference on Computational Intelligence and Security (CIS)

PB - IEEE

T2 - 2025 21st International Conference on Computational Intelligence and Security, CIS 2025

Y2 - 12 December 2025 through 15 December 2025

ER -

An AI Education Framework Based on LLM-Knowledge Graph for Personalized Learning

Abstract

Conference

User-Defined Keywords

Access to Document

Other files and links

Fingerprint

Cite this