Elamur Research⌕
☰
home / papers / 2026 / CogNet-KG: Empowering Tutoring Dialogues with a Cognitively-Structured Knowledge Graph for STEM Learning

CogNet-KG: Empowering Tutoring Dialogues with a Cognitively-Structured Knowledge Graph for STEM Learning

Paper · 2026 · Ding Yu, Yu Lu, Tengju Li et al.

Summary

Educational knowledge graphs used in intelligent tutoring systems usually link concepts only through generic relations such as similarity or prerequisite. They tend to cover one subject at a time and rarely connect to explicit learning objectives, so they offer little help in guiding interactive teaching. The authors address this with CogNet-KG, a graph covering roughly 500 core secondary-school concepts in five STEM subjects. It was built from the expert-annotated SLP dataset. A pipeline of LLM agents reads each learning objective, finds the concepts it mentions, maps them to the knowledge base, and infers typed cognitive relations, which are then reviewed. Each relation is tied to a concrete objective. The graph then drives the generation of CogDialogue-QA, a set of simulated diagnostic tutoring dialogues in which the teacher's next move (remediation, consolidation or extension) depends on the student's answer and the graph structure. Fine-tuning a base model on these dialogues yields CogTutor-LM. In evaluation, it produced more coherent and pedagogically appropriate guidance than general LLMs and a RAG baseline. This suggests that structured cognitive knowledge can improve LLM tutors.

Key points

  • CogNet-KG spans nearly 500 concepts across Mathematics, Physics, Chemistry, Biology and Geography. It uses 9 relation types (7 directed, 2 undirected), such as transfer, apply and compare, which were designed with subject experts and drawn from Bloom's Taxonomy.
  • The construction pipeline has three agents:
    • Knowledge Recognizer (Agent_A) extracts the concepts mentioned in each learning objective.
    • Knowledge Selector (Agent_B) matches them to the knowledge base using embedding retrieval of the top-10 candidates by cosine similarity.
    • Relation Analyst (Agent_C) proposes relation triples, which then go through a review stage.
  • CogTutor-LM showed the largest gains in Expression, Coherence, Constructiveness and Guidance. Its improvements in Diagnosis and Motivation were small. GPT-4o was generally the second-best system.
  • The RAG baseline underperformed because similarity-based retrieval cannot choose objectives according to cognitive structure.
  • The datasets are released at github.com/KCAIED/CogNet-KG. The authors list three limitations:
    • The work covers STEM subjects only.
    • The cross-grade and cross-subject links are not exploited in the dialogue generation or the fine-tuning.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Institutions

🎓Beijing Normal UniversityBeijing · China22 works · 3,177 citations

 Fields, topics and concepts

📚 Shelves 5The field of science this work belongs to (OpenAlex fields).Computer Science · 18,614 worksArtificial Intelligence · 17,965 worksSocial Sciences · 1,954 worksInformation Systems · 1,730 worksDemography · 9 works

🧭 Topics 1Research subjects and languages the work is about.Dialogue systems · 654 works

Cite

Cite as Ding Yu, Yu Lu, Tengju Li, Shasha Xiong and Shengquan Yu. 2026. CogNet-KG: Empowering Tutoring Dialogues with a Cognitively-Structured Knowledge Graph for STEM Learning. Findings of the Association for Computational Linguistics: ACL 2026.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.findings-acl.639

@inproceedings{yu2026cognet,
  title = {{CogNet-KG: Empowering Tutoring Dialogues with a Cognitively-Structured Knowledge Graph for STEM Learning}},
  author = {Ding Yu, Yu Lu, Tengju Li, Shasha Xiong and Shengquan Yu},
  booktitle = {Findings of the Association for Computational Linguistics: ACL 2026},
  year = {2026},
  doi = {10.18653/v1/2026.findings-acl.639},
  url = {https://doi.org/10.18653/v1/2026.findings-acl.639}
}

References

34 references, 6 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Garima Agrawal, Kuntal Pal, Yuli Deng, Huan Liu, and Ying-Chih Chen. 2024. CyberQ: Generating Questions and Answers for Cybersecurity Education Using Knowledge Graph-Augmented LLMs. Proceedings of the AAAI Conference on Artificial Intelligence, 38(21):23164–23172. find
  2. Omaima Almatrafi and Aditya Johri. 2025. Leveraging generative ai for course learning outcome categorization using bloom’s taxonomy. Computers and Education: Artificial Intelligence, 8:100404. 2025. From objectives to questions: A planning-based framework for educational mathematical question genera… find
  3. Yuhao Dan, Zhikai Lei, Yiyang Gu, Yong Li, Jianghao Yin, Jiaju Lin, Linhao Ye, Zhiyan Tie, Yougen Zhou, Yilei Wang, Aimin Zhou, Ze Zhou, Qin Chen, Jie Zhou, Liang He, and Xipeng Qiu. 2023. EduChat: A Large-Scale Language Model-based Chatbot System for Intelligent Education.
  4. Fu-Rong Dang, Jin-Tao Tang, Kun-Yuan Pang, Ting Wang, Sha-Sha Li, and Xiao Li. 2021. Constructing an Educational Knowledge Graph with Concepts Linked to Wikipedia. Journal of Computer Science and Technology, 36(5):1200–1211. find
  5. Lynn S. Fuchs, Douglas Fuchs, Karin Prentice, Carol L. Hamlett, Robin Finelli, and Susan J. Courey. 2004. Enhancing mathematical problem solving among third-grade students with schema-based instruction. Journal of Educational Psychology, 96(4):635–647. find
  6. Weibo Gao, Qi Liu, Zhenya Huang, Yu Yin, Haoyang Bi, Mu-Chun Wang, Jianhui Ma, Shijin Wang, and Yu Su. 2021. RCD: Relation map driven cognitive diagnosis for intelligent education systems. In Proceedings of the 44th International ACM SIGIR Conference on Research and Development in Information Retrie… find
  7. Team GLM, :, Aohan Zeng, Bin Xu, Bowen Wang, Chenhui Zhang, Da Yin, Dan Zhang, Diego Rojas, Guanyu Feng, Hanlin Zhao, Hanyu Lai, Hao Yu, Hongning Wang, Jiadai Sun, Jiajie Zhang, Jiale Cheng, Jiayi Gui, Jie Tang, and 40 others. 2024. Chatglm: A family of large language models from glm-130b to glm-4 a…
  8. Ed Hammond, Jennifer. 2001. Scaffolding: Teaching and Learning in Language and Literacy Education. Scaffolding: Teaching and Learning in Language and Literacy Education. find
  9. R. E. Haskell. 2001. Transfer of learning: Cognition, instruction, and reasoning. Academic Press. find
  10. Lorin W Anderson, David R Krathwohl, Peter W Airasian, Kathleen A Cruikshank, Richard E Mayer, Paul R Pintrich, James Raths, and Merlin C Wittrock. 2001. A Taxonomy for Learning, Teaching, and Assessing: A Revision of Bloom’s Taxonomy of Educational Objectives. Longman, New York, NY. find
  11. Nan Li, Qiang Shen, Rui Song, Yang Chi, and Hao Xu. 2022. MEduKG: A deep-learning-based approach for multi-modal educational knowledge graph construction. Information-an International Interdisciplinary Journal, 13(91). find
  12. Penghe Chen, Yu Lu, Vincent W. Zheng, Xiyang Chen, and Boda Yang. 2018. KnowEdu: A System to Construct Knowledge Graph for Education. IEEE Access, 6:31553–31563. find
  13. Yu Lu, Penghe Chen, Yang Pian, and Vincent W. Zheng. 2022. CMKT: Concept Map Driven Knowledge Tracing. IEEE Transactions on Learning Technologies, 15(4):467–480. find
  14. Cheng Cheng, Zhenya Huang, GuanHao Zhao, Yuxiang Guo, Xin Lin, Jinze Wu, Xin Li, and Shijin Wang. find
  15. Yu Lu, Yang Pian, Ziding Shen, Penghe Chen, and Xiaoqing Li. 2021. SLP: A Multi-Dimensional and Consecutive Dataset from K-12 Education. In 29TH INTERNATIONAL CONFERENCE ON COMPUT-ERS IN EDUCATION (ICCE 2021), VOL I, pages 261–266, NO 300, JUNGDA RD, JHONGLI DIS-TRICT, TAOYUAN CITY, 320, TAIWAN. ASI… find
  16. Rui Lv, Qi Liu, Weibo Gao, Jiatong Li, Kai Zhang, and Shiwei Tong. 2025a. ReAL: How can LLMs simulate the real teacher? Retrieval-enhanced agent for adaptive learning. In Findings of the Association for Computational Linguistics: EMNLP 2025, pages 168–181, Suzhou, China. Association for Computationa… find
  17. Rui Lv, Qi Liu, Weibo Gao, Jiatong Li, Kai Zhang, and Shiwei Tong. 2025b. ReAL: How can LLMs simulate the real teacher? Retrieval-enhanced agent for adaptive learning. In Findings of the Association for Computational Linguistics: EMNLP 2025, pages 168–181, Suzhou, China. Association for Computationa… find
  18. Long Ouyang, Jeff Wu, Xu Jiang, Diogo Almeida, Carroll L. Wainwright, Pamela Mishkin, Chong Zhang, Sandhini Agarwal, Katarina Slama, Alex Ray, John Schulman, Jacob Hilton, Fraser Kelton, Luke Miller, Maddie Simens, Amanda Askell, Peter Welinder, Paul Christiano, Jan Leike, and Ryan Lowe. 2022. Train…
  19. Xian Peng, Pan Yuan, Dong Li, Junlong Cheng, Qin Fang, and Zhi Liu. 2025. KELE: A multi-agent framework for structured socratic teaching with large language models. In Findings of the Association for Computational Linguistics: EMNLP 2025, pages 16342–16362, Suzhou, China. Association for Computation… find
  20. Changyong Qi, Yuang Wei, Haoxin Xu, Longwei Zheng, Peiji Chen, and Xiaoqing Gu. 2025. TMATH a dataset for evaluating large language models in generating educational hints for math word problems. In Proceedings of the 31st International Conference on Computational Linguistics, pages 5082–5093, Abu Dh… find
  21. Kechen Qu, Kam Cheong Li, Billy T. M. Wong, Manfred M. F. Wu, and Mengjin Liu. 2024. A survey of knowledge graph approaches and applications in education. Electronics, 13(2537). find
  22. Qwen, :, An Yang, Baosong Yang, Beichen Zhang, Binyuan Hui, Bo Zheng, Bowen Yu, Chengyuan Li, Dayiheng Liu, Fei Huang, Haoran Wei, Huan Lin, Jian Yang, Jianhong Tu, Jianwei Zhang, Jianxin Yang, Jiaxi Yang, Jingren Zhou, and 25 others. 2025. Qwen2.5 technical report. Preprint, arXiv:2412.15115. find
  23. Shashank Sonkar, Naiming Liu, Debshila Mallick, and Richard Baraniuk. 2023. CLASS: A design framework for building intelligent tutoring systems based on learning science principles. In Findings of the Association for Computational Linguistics: EMNLP 2023, pages 1941–1961, Singapore. Association for… find
  24. Ying Su and Yong Zhang. 2020. Automatic Construction of Subject Knowledge Graph based on Educational Big Data. In Proceedings of the 2020 3rd International Conference on Big Data and Education, ICBDE ’20, pages 30–36, New York, NY, USA. Association for Computing Machinery. find
  25. Hugo Touvron, Thibaut Lavril, Gautier Izacard, Xavier Martinet, Marie-Anne Lachaux, Timothée Lacroix, Baptiste Rozière, Naman Goyal, Eric Hambro, Faisal Azhar, Aurelien Rodriguez, Armand Joulin, Edouard Grave, and Guillaume Lample. 2023. Llama: Open and efficient foundation language models. Preprint…
  26. L. S. Vygotsky. 1978. Mind in Society: Development of Higher Psychological Processes. Harvard University Press. find
  27. Haochun Wang, Sendong Zhao, Zewen Qiang, Zijian Li, Chi Liu, Nuwa Xi, Yanrui Du, Bing Qin, and Ting Liu. 2025. Knowledge-tuning large language models with structured medical knowledge bases for trustworthy response generation in chinese. ACM Trans. Knowl. Discov. Data, 19(2). find
  28. Xin Wei, Wenrui Han, Shiyun Sun, Junhao Shan, and Liang Zhou. 2025. Educational psychology-empowered personalized learning path generation strategy. IEEE Transactions on Learning Technologies, 18:741–756. find
  29. Lixiang Yan, Lele Sha, Linxuan Zhao, Yuheng Li, Roberto Martinez-Maldonado, Guanliang Chen, Xinyu Li, Yueqiao Jin, and Dragan Gašević. 2024. Practical and ethical challenges of large language models in education: A systematic scoping review. British Journal of Educational Technology, 55(1):90– 112. find
  30. An Yang, Anfeng Li, Baosong Yang, Beichen Zhang, Binyuan Hui, Bo Zheng, Bowen Yu, Chang Gao, Chengen Huang, Chenxu Lv, Chujie Zheng, Dayiheng Liu, Fan Zhou, Fei Huang, Feng Hu, Hao Ge, Haoran Wei, Huan Lin, Jialong Tang, and 41 others. 2025. Qwen3 technical report. Preprint, arXiv:2505.09388.
  31. Weihao You, Pengcheng Wang, Changlong Li, Zhilong Ji, and Jinfeng Bai. 2024. CK12: A Rounded K12 Knowledge Graph Based Benchmark for Chinese Holistic Cognition Evaluation. Proceedings of the AAAI Conference on Artificial Intelligence, 38(17):19431–19439. find
  32. Xueqiao Zhang, Chao Zhang, Jianwen Sun, Jun Xiao, Yi Yang, and Yawei Luo. 2025. EduPlanner: LLM-based multiagent systems for customized and intelligent instructional design. IEEE Transactions on Learning Technologies, 18:416–427. find
  33. Bowen Zhao, Jiuding Sun, Bin Xu, Xingyu Lu, Yuchen Li, Jifan Yu, Minghui Liu, Tingjian Zhang, Qiuyang Chen, Hanming Li, Lei Hou, and Juanzi Li. 2022. EDUKG: A heterogeneous sustainable K-12 educational knowledge graph. Preprint, arXiv:2210.12228. — [arXiv:2210.12228](https://arxiv.org/abs/2210.12228) find
  34. Wayne Xin Zhao, Kun Zhou, Junyi Li, Tianyi Tang, Xiaolei Wang, Yupeng Hou, Yingqian Min, Beichen Zhang, Junjie Zhang, Zican Dong, Yifan Du, Chen Yang, Yushuo Chen, Zhipeng Chen, Jinhao Jiang, Ruiyang Ren, Yifan Li, Xinyu Tang, Zikang Liu, and 3 others. 2023. A Survey of Large Language Models.

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/yu-2026-cognet-kg-empowering-tutoring-dialogues/