EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Bidirectional fusion heterogeneous graph networks for semi-supervised Bitcoin transaction anomaly detection in dynamic transaction graphs

Bo Xiao, Wei Yin iD

PLOS One · 08.06.2026 · 21 (6) · e0351051

DOI: 10.1371/journal.pone.0351051 ↗

Kurz gefasst

Fachartikel in der Zeitschrift PLOS One, online erschienen am 08.06.2026. Verfasst von Bo Xiao und 1 weiteren Autor:innen. Beteiligte Einrichtung: School of Economics and Management, Southeast University, Nanjing, China. Hauptthemen laut MEDLINE: Blockchain, Computer Security, Graph Neural Networks. Die Arbeit nennt 62 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Social Science Fund of China.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Detecting anomalies in the Bitcoin transaction network is critical for ensuring blockchain security and stability. The network’s heterogeneous structure and dynamic nature, coupled with scarce labeled anomalies, pose significant challenges for traditional graph-based methods. To address these, we propose Bidirectional Fusion Heterogeneous Graph Network (BF-HGN), a semi- dynamic supervised model for Bitcoin transaction anomaly detection task. BF-HGN designs multi-type feature embedding and alignment strategies to effectively unify features across heterogeneous transaction–address nodes. A bidirectional temporal fusion mechanism is proposed to capture long-range temporal dependencies that unidirectional models often miss. To alleviate class imbalance and limited annotations, a Class-balanced Classifier (CBC) combined with Adjacency Adaptation (AA) and Adaptive Feature Space Regulation (AFSR) losses is proposed to generate pseudo-anomalous nodes closely resembling real anomalies, improving discrimination boundaries. Experiments on the Elliptic++ dataset demonstrate that BF-HGN outperforms existing methods, achieving F1 scores of 0.6301 and 0.5784 for transaction and address nodes, respectively, establishing a new benchmark for Bitcoin transaction anomaly detection.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

BlockchainComputer SecurityGraph Neural Networks

Autor:innen und Institutionen

  • School of Economics and Management, Southeast University, Nanjing, China.
  • The Laboratory of Philosophy and Social Sciences at Universities in Jiangsu Province-Fintech and Big Data Laboratory of Southeast University, Nanjing, China.

ORCID-Profile: Wei Yin

Förderung

  • National Social Science Fund of China

Zitieren

APA

Xiao, B., & Yin, W. (2026). Bidirectional fusion heterogeneous graph networks for semi-supervised Bitcoin transaction anomaly detection in dynamic transaction graphs. PLOS One, 21(6), e0351051. https://doi.org/10.1371/journal.pone.0351051

Vancouver

Xiao B, Yin W. Bidirectional fusion heterogeneous graph networks for semi-supervised Bitcoin transaction anomaly detection in dynamic transaction graphs. PLOS One. 2026;21(6):e0351051. doi:10.1371/journal.pone.0351051

BibTeX
@article{Xiao2026,
  author = {Xiao, Bo and Yin, Wei},
  title = {Bidirectional fusion heterogeneous graph networks for semi-supervised Bitcoin transaction anomaly detection in dynamic transaction graphs},
  journal = {PLOS One},
  year = {2026},
  volume = {21},
  number = {6},
  pages = {e0351051},
  doi = {10.1371/journal.pone.0351051}
}
Kostenlos starten →