EUVIMEDEUVIMED entdecken
Crossref

Cybercrime classification using keyword-based knowledge graphs and graph neural networks

Heehun Jeong, Daeun Lee, Sunho Kim, Susik Yoon, Dong-joo Kim

Journal of Data Forensics Research · 31.12.2025 · 2 (2) · 175-195

DOI: 10.12972/jdfr.2025.2.2.4 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Data Forensics Research, online erschienen am 31.12.2025. Verfasst von Heehun Jeong und 4 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

With the spread of digital financial transactions and online communication channels, cybercrime involving Telegramand websites has been constantly increasing. However, clear limitations exist in manually analyzing and classifying vast text-based open-source intelligence (OSINT) dataaccordingly, the need for an automated classification system is growing. To solve this problem, this study proposes an integrated framework for automatically classifying crime conditions and types by structuring Telegram channel messages and website HTML into keyword-based knowledge graphs and applying graph neural networks (GNNs). Structured knowledge graphs are used to learn graph convolutional networks (GCNs) for multiclass crime type classification of Telegram channels and binary crime/normal classification of websites. As a result of the experiment, the GCN model showed overall superior performance compared with comparative models such as logistic regression, multilayer perceptron (MLP), and XGBoost and achieved an F1 score of approximately 0.80 in Telegram crime type classification and approximately 0.92 in website crime detection. Taken together, this study proposes the basis for automating cybercrime detection and classification by connecting text-based OSINT to a knowledge graph–GNN workflow and expanding it to multimodal and heterogeneous graph-based threat intelligence systems in the future.

Zitieren

APA

Jeong, H., Lee, D., Kim, S., Yoon, S., & Kim, D. (2025). Cybercrime classification using keyword-based knowledge graphs and graph neural networks. Journal of Data Forensics Research, 2(2), 175-195. https://doi.org/10.12972/jdfr.2025.2.2.4

Vancouver

Jeong H, Lee D, Kim S, Yoon S, Kim D. Cybercrime classification using keyword-based knowledge graphs and graph neural networks. Journal of Data Forensics Research. 2025;2(2):175-195. doi:10.12972/jdfr.2025.2.2.4

BibTeX
@article{Jeong2025,
  author = {Jeong, Heehun and Lee, Daeun and Kim, Sunho and Yoon, Susik and Kim, Dong-joo},
  title = {Cybercrime classification using keyword-based knowledge graphs and graph neural networks},
  journal = {Journal of Data Forensics Research},
  year = {2025},
  volume = {2},
  number = {2},
  pages = {175--195},
  doi = {10.12972/jdfr.2025.2.2.4}
}
Kostenlos starten →