Study on digital financial fraud risk identification based on heterogeneous graph convolutional attention network

Yang Jin

International Journal of Networking and Virtual Organisations, 2025, vol. 32, issue 1/2/3/4, 203-218

Abstract: To enhance the accuracy of digital financial fraud risk identification and reduce the identification time, this paper introduces a digital financial fraud risk identification method utilising a heterogeneous graph convolutional attention network. Initially, financial business data is gathered and processed for data imbalance using generative adversarial networks. Subsequently, extreme learning machines are employed to extract spatially correlated features among various transactions. Following this, a robust graph convolutional attention network is constructed, a risk identification function is designed, and ultimately, the fraud data is fed into the graph convolutional neural network for training. The output data is categorised by transaction type to ascertain the presence of digital financial fraud risk. The results indicate that our method achieves a recognition accuracy exceeding 96%, with time consumption not surpassing 8.5 s, demonstrating that our method exhibits excellent recognition performance.

Keywords: heterogeneous graph convolutional attention network; LSTM training; digital finance; fraud risk; extreme learning machine. (search for similar items in EconPapers)
Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
http://www.inderscience.com/link.php?id=145376 (text/html)
Access to full text is restricted to subscribers.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:ids:ijnvor:v:32:y:2025:i:1/2/3/4:p:203-218

Access Statistics for this article

More articles in International Journal of Networking and Virtual Organisations from Inderscience Enterprises Ltd
Bibliographic data for series maintained by Sarah Parker ().