Design of Financial Anti-fraud Application System Driven by Blockchain and Federated Learning
Quan Yizhuo, Hu Chenlu · DEVELOPMENT ECONOMICS OF CHINA · 2025
With the rapid development of financial technology, financial fraud methods are constantly evolving and upgrading, bringing severe challenges to the security and stability of the financial industry. Traditional anti-fraud methods are difficult to meet the current needs of financial risk prevention and control due to problems such as data silos and privacy protection. To this end, this paper proposes and designs a financial anti-fraud application system architecture that integrates blockchain and federated learning technology. Through federated learning, the system realizes joint modeling of multiple institutions without sharing original data, thereby improving the accuracy and robustness of fraud detection models; at the same time, combined with the tamper-proof and traceable characteristics of blockchain, it ensures the transparency and credibility of model training and parameter interaction processes, further enhancing the security and compliance of the system. This paper analyzes the functional architecture and key technical implementation of the system in detail, including model construction, privacy protection mechanism, smart contract design, etc., designs specific interaction processes, and conducts prototype development and experimental verification. The experimental results show that the proposed system can effectively improve the accuracy and real-time response capabilities of financial fraud detection while ensuring data privacy and security. This study provides new ideas and practical references for the development of financial anti-fraud technology under the background of multi-institutional collaboration.