Main article

Nur Aisyah Rahman
Faculty of Computing and Informatics, Universiti Malaysia Sabah, Kota Kinabalu, Malaysia
Daniel Lim Wei Shen
Department of Information Systems, Universiti Tunku Abdul Rahman, Kampar, Malaysia
Faridah Mohd Yusof*
Faculty of Industrial Management, Universiti Malaysia Pahang Al-Sultan Abdullah, Pahang, Malaysia
faridah.yusof@umpsa.edu.my

Abstract

Blockchain analytics has become a central instrument of FinTech risk management because public ledgers preserve transaction traces while also enabling pseudonymous movement of value across jurisdictions, exchanges, mixers, bridges, and decentralized applications. Existing graph-based anomaly detection studies have shown that dynamic heterogeneous graph learning can improve the identification of suspicious transactions and addresses under limited labels, but the innovation challenge is not only technical. Surveillance systems influence customer screening, compliance workload, institutional trust, financial inclusion, and accountability. This article develops a sociotechnical framework for blockchain transaction surveillance that integrates graph-based anomaly detection with responsible FinTech innovation. Building on the research direction of dynamic heterogeneous Bitcoin transaction graphs, the paper proposes a Blockchain Fraud Graph Surveillance and Trust framework (BFG-ST) that aligns four layers: data governance, graph representation, anomaly scoring, and human-centered institutional oversight. A simulated evaluation design based on the Elliptic++ task structure is used to illustrate how transaction and address nodes, temporal slices, class imbalance, and analyst feedback can be combined in a semi-supervised monitoring pipeline. The results suggest that responsible graph surveillance should not be assessed only by F1-score or recall. It should also be evaluated through explainability coverage, alert burden, review fairness, computational proportionality, and policy traceability. The study contributes a governance-oriented analytics architecture for responsible blockchain intelligence and provides design principles for regulators, exchanges, compliance teams, and FinTech innovators seeking to use graph learning without creating opaque or excessive surveillance.

Article details

How to Cite

Rahman, N. A., Wei Shen, D. L., & Yusof, F. M. (2025). Blockchain Transaction Surveillance and Responsible FinTech Innovation: A Sociotechnical Framework for Graph-Based Anomaly Detection. Journal of Technology Innovation and Society, 3(3), 30-46. https://doi.org/10.63646/jtis.2025.030302