Education, Science, Technology, Innovation and Life
Open Access
Sign In
  1. Home
  2. Journals
  3. Advances in Computer, Signals and Systems
  4. Vol 10, Issue 2, 2026

Research on Early Warning Method for Academic Misconduct in Universities Based on Knowledge Graph and Graph Neural Network

Download as PDF

DOI: 10.23977/acss.2026.100219 | Downloads: 5 | Views: 59

Author(s)

Yang Lei 1

Affiliation(s)

1 Yunnan Open University, Kunming, China

Corresponding Author

Yang Lei

ABSTRACT

Against the complex situation where academic misconduct is growing more concealed, better organized and enabled by upgraded technical means, the traditional governance model relying on manual review and ex post punishment can hardly cope with current challenges. The governance of research integrity is shifting from passive response to active prevention and control. This paper proposes a knowledge graph construction and Graph Neural Network (GNN) early warning method for university research integrity governance. First, a research integrity domain ontology covering core entities such as researchers, projects, research outputs, institutions and journals is designed, and a knowledge graph is built upon multi source heterogeneous research data. Second, given the real world constraint of sparse labelled data for academic misconduct, an anomaly detection early warning model is constructed combining the Relational Graph Convolutional Network (RGCN) and Graph Attention Network (GAT). Within a graph auto encoder framework, deep representations of normal research behaviours are learned in an unsupervised manner, and abnormal patterns such as improper citation and suspicious collaboration are identified via reconstruction error. Finally, a prototype early warning system is implemented accordingly, and the human machine collaborative governance path of "intelligent discovery — manual verification — targeted intervention" is discussed. This study provides references for technical integration and application in research integrity governance for university research management departments.

KEYWORDS

Academic misconduct; Knowledge graph; Graph neural network; Anomaly detection; Early warning model

CITE THIS PAPER

Yang Lei. Research on Early Warning Method for Academic Misconduct in Universities Based on Knowledge Graph and Graph Neural Network. Advances in Computer, Signals and Systems (2026). Vol. 10, No. 2, 167-174. DOI: http://dx.doi.org/10.23977/acss.2026.100219.

REFERENCES

[1] Wang Shao. ChatGPT and the Governance of Academic Misconduct: Challenges and Countermeasures[J]. Science & Technology Progress and Policy, 2023, 40(14): 120 127.
[2] Hu Guohui, Wang Zongshui, Sun Zhuo, et al. Retraction Reason Analysis and Academic Deviance Governance Based on Social Network Analysis[J]. Chinese Journal of Scientific and Technical Periodicals, 2024, 35(12): 1744 1756.
[3] Yuan Zihan, Liu Yi. Ethical Issues in Retracted COVID 19 Related Papers and Their Governance Paths[J]. Medicine and Philosophy, 2024, 45(3): 22 26.
[4] Chen Liuyuan, Zhang Xue. Research Integrity Governance in Universities: Challenges and Countermeasures[J]. Journal of Xinyang Normal University (Philosophy and Social Sciences Edition), 2025, 45(3): 79 83.
[5] Zhang Huaibo, Liu Xiaona, Liu Xin, et al. A Microblog Social Bot Detection Method Based on Multi heterogeneous Networks with Fused Features[J]. Journal of Chinese Information Processing, 2025, 39(1): 133 143.
[6] Li Ruoyu. Reconstruction of the External Governance System for Research Integrity from the Perspective of Holistic Governance[J]. Science & Technology Progress and Policy, 2025, 42(3): 85 96.
[7] Liu Shengli, Pan Yuntao. Measurement Dilemma and Irrational Risks of Research Integrity — Academic Reflection on the Technical Approach of External Norms for Research Integrity[J]. Science & Technology Progress and Policy, 2023, 40(20): 21 29.
[8] Wang Ying, Qian Li, Xie Jing, et al. Research on Construction Model and Methods of Knowledge Graph for Scientific and Technological Big Data[J]. Data Analysis and Knowledge Discovery, 2019, 3(1): 15 26.
[9] Zhou Yuanchun, Wang Weijun, Qiao Ziyue, et al. Review on Construction Methods and Applications of Knowledge Graph for Scientific Big Data[J]. Scientia Sinica(Informationis), 2020, 50(7): 957 987.
[10] KIM H, LEE B, SHIN W. Graph anomaly detection with graph neural networks: current status and challenges [J]. IEEE Access,2022.
[11] Guo Xiaoliang, Zhang Lu, Yin Shuying, et al. Exploration and Construction of Risk grading Early warning System for Academic Misconduct in Publishing Field[J]. Chinese Journal of Scientific and Technical Periodicals, 2025, 36(8): 1045 1057.
[12] Pan Kexu. Multi level Business Process Anomaly Detection Based on Transformer[D]. Huainan: Anhui University of Science and Technology, 2026.DOI:10.26918/d.cnki.ghngc.2026.001553.
[13] ZHAO L, SAWLANI S, SRINIVASAN A. Graph anomaly detection with unsupervised gnns [EB/OL]. arXiv:2210.09535,2022.

Subscription
E-Mail Alert
Most popular papers
Call for Special Issue
Publication Ethics & OA Statement
Plagiarism Policy
Contact Us
Downloads: 51671
Visits: 1356841

Sponsors, Associates, and Links

All published work is licensed under a Creative Commons Attribution 4.0 International License.

Copyright © 2016 - 2031 Clausius Scientific Press Inc. All Rights Reserved.